KR101944288B1 - Laminated, leak-resistant chemical processors, methods of making, and methods of operating - Google Patents

Abstract

The present invention provides a method of manufacturing a laminated device (particularly a microchannel device) welded together by plates. Unlike a conventional microchannel device, the laminated device of the present invention can be performed without brazing or diffusion bonding, which can provide significant advantages in manufacturing. Features such as expansion joints and external weld supports are also described. A method of carrying out a unit operation in a laminated type apparatus and a laminated type apparatus is also described.

Description

TECHNICAL FIELD [0001] The present invention relates to a laminate type leakage resistance chemical processor, a manufacturing method thereof, and a method of manufacturing the same.

This application claims priority to U.S. Provisional Application No. 61 / 394,328, filed October 18, 2010, and 61 / 441,276, filed February 9, 2011,

The conventional idea in microchannel technology is that the optimal heat exchange in the microchannel heat exchanger can only be obtained by diffusion bonding and / or brazing. This method relies on the formation of adjacent metal interfaces between the layers. This adjacent interface is believed to be necessary to eliminate the thermal contact resistance between the layers and to remove high levels of heat from the heat dissipation reaction to the heat rejection chamber or to add heat to the heat dissipation.

Brazing requires the addition of an interlayer material that melts at a temperature below the melting temperature of the constituent material. The intermediate layer becomes liquid during the diffusion brazing or brazing process. The liquid interlayer flows to fill the gaps or gaps so that the materials are bonded together. Further, since the intermediate layer flows and diffuses, the material from the intermediate layer can be diffused into the base material, and the material from the base material can be diffused into the intermediate layer.

As diffusion proceeds, the local composition of the intermediate material is changed. Further, when the temperature is lowered after the maximum temperature is reached, the liquid intermediate layer solidifies and fills the gap between the two base material layers. The solidification can proceed by temperature or composition. In the latter case, a melting point depressant such as phosphorous acid or boron may be added so that the intermediate layer melts at a lower temperature than the parent material. In a similar example, the diffusion joining device also forms a tight thermal contact between the heat transfer layers.

In a first embodiment, the present invention provides a laminated device comprising a bottom sheet and an upper sheet, wherein a gap is disposed between the top sheet surface and the bottom sheet surface, wherein the top and bottom sheets are part of a subassembly , Providing a thermally conductive pin insert having a height that is at least 1% (preferably at least 2%, and in some embodiments at least 1 to 10%) greater than the gap height; Placing the pin insert in the gap, and pressing the sheet such that the pin insert is deformed to fit within the gap.

The insert may or may not have a catalyst coating. In some preferred embodiments, the final device has an irregular configuration in which, for example, the crushed walls are not all bent in the same direction or are deformed in an irregular manner that does not bend in an alternating direction. In some embodiments, the fins are bent in the same direction (greater than 50%, preferably greater than 80%). The compression pin does not have the regular shape as it exists in the pleated sheet.

In some preferred embodiments, the bottom and / or top sheet surface has a weld line projecting from the surface, and in some preferred embodiments the sheet has a continuous length of at least 50% (preferably at least 80%) of the length or width of the sheet And at least one side of the fin insert is at least partially (preferably completely) laid on at least one weld line. Preferably, the direction of flow through the insert is perpendicular to the weld line. Surprisingly, it has been found that good heat transfer is obtained by the presence of these weld lines even when the weld line is interposed between the heat exchanger and the reaction chamber (i.e., the chamber containing the press-fit addition catalyst) The pin insert is a tack welded to the top or bottom surface and in a more preferred embodiment the pin insert is held in place by the press-fit and is not bonded to either side by welding. Preferably, the fins are disposed in a microchannel, and / or the adjacent heat exchanger comprises a microchannel.

For all of the methods described herein, the present invention also encompasses devices formed by such methods. It also includes an operation of the apparatus formed by the method for performing one or more unit operations. Preferably, the apparatus described herein is a chemical process suitable for performing one or more unit operations.

In another aspect, the present invention provides an apparatus comprising a device having a fluid inlet connected to a process chamber and a fluid outlet connected to the process chamber, and a heat exchanger in thermal contact with the process chamber, wherein the process chamber is disposed in a gap and a gap, And a thermally conductive pin insert in contact with the bottom, the pin insert having an irregular shape caused by at least 1% compression of the pin insert in the gap. In such an apparatus, the height of the pin insert is at least 1%, preferably at least 2%, in some embodiments at least 5%, and in other embodiments from 1 to 10% of the pin height (height in the same direction as the stack height) The irregular shape is generated by compression of the sheet on either side of the gap such that it is deformed by a range.

Such a device may include any feature described herein. By way of example, in some preferred embodiments, the insert has no catalyst coating, and in some preferred embodiments, the catalyst particles are in the gap. The conductive pin insert is a fin that transfers heat from the process occurring within the fin section to the adjacent heat exchange layer. Preferably, the fin insert is made of a material having a higher thermal conductivity than the constituent material of the heat transfer layer. In some embodiments, the conductive pin insert has a thermal conductivity greater than 10 times, and more preferably greater than 100 times, the effective thermal conductivity of the catalyst disposed within the conductive pin structure. For most materials, the thermal conductivity is known and, if not, can be measured using standard ASTM methods.

In another aspect, the present invention is directed to a process for preparing a composition comprising the steps of passing at least one reactant into a process chamber of the apparatus of claim 2, performing a process within the process chamber, and simultaneously exchanging heat between the process chamber and the heat exchanger Provides a chemical reaction method.

In yet another aspect, the present invention provides a method of manufacturing a welded edge, comprising providing a first subassembly or first sheet and a second subassembly or second sheet, Depositing the welded single sheet or combination layer with at least one layer or sheet; and depositing a layer or layers laminated to form a laminate-like device, Wherein the first subassembly or first sheet comprises a first parallel array of channels and the second subassembly or second sheet comprises a second parallel array of channels, There is no intersection between the subassembly or first sheet and second subassembly or second sheet.

The first and second arrays of parallel channels may share a common header and / or footer, although there is no cross channel in which the channels in the first and second sheets are mixed. In one embodiment of this method, the sheet (or a subassembly comprising a plurality of sheets) is cut into four pieces and the four pieces are welded together back along their edges - this cuts the sheet It is a semi-intuitive process of rejoining along the edge that was cut before. Nonetheless, it has been found that this process significantly reduces warpage, making it possible to produce superior laminate-like products. In some broad aspects of this method, the first and second sheets (or subassemblies) can be obtained separately without being cut from a common piece. Preferably, the stack layer is of the same width and length (within 5% of length and width in some embodiments, and within 1% of length and width in some embodiments). The cutting (and re-welding) should be performed parallel to the channel length, and there is no risk of forming a blocking or discontinuous channel in this direction.

In some preferred embodiments, the aspect ratio of the first subassembly at width: height, and length: height is greater than 2, and preferably greater than 10. In some preferred embodiments, the aspect ratio of the first sheet or first subassembly at width: height, or length: width is at least 1.5, more preferably at least 2, and in some embodiments at least 4. Height refers to lamination height, width and length are mutually perpendicular, and length is the direction of fluid flow along the sheet. Preferably, the first subassembly or first sheet has at least five first parallel channels (preferably microchannels) and the second subassembly or second sheet has at least five second parallel channels, Channel), and the first and second subassemblies or sheets are joined along the edge such that the first parallel channel in the sheet or subassembly is parallel to the second parallel channel, preferably a single manifold is disposed between the first and second parallel It functions as both channels. Preferably, the first and second subassemblies or sheets are bonded such that the adjacent parallel channels in the first and second assemblies or sheets are oriented in a transverse direction (the length parallel to the net direction of flow through the device) Cm, and more preferably, within 1 cm. The final device includes a sheet having a welded portion which is believed to connect the segments of the sheet. In some preferred embodiments, the step of planarizing one or more (preferably all) subassemblies, when the aspect ratio of width: height or length: height of the first sheet or first subassembly is greater than 1, Before welding together.

By all the methods described herein, the present invention encompasses products having results from such methods.

The present invention has a first layer and a second layer having length and width dimensions, the first layer comprising a first section having a first plurality of parallel channels and a second section having a second plurality of parallel channels Wherein the first and second plurality of channels are parallel and the first and second sections are joined together by a junction, the junction is parallel to the parallel channel, and the second layer comprises a laminate of a first layer Lt; / RTI > In some preferred embodiments, the first layer is a subassembly and the second layer is a subassembly, and the layers are joined by welding along the periphery of the layer.

In another aspect, the present invention provides a pre-cambering method for fabricating a flatter subassembly. Such a laminate-type device manufacturing method includes the steps of providing a metal sheet, deforming the metal sheet, and bonding the deformed metal sheet to the laminate-type device.

In a preferred embodiment, the metal sheet is deformed into an arcuate shape having a curvature of at least 10 degrees from the flat portion, and in some embodiments from 10 to 80 degrees from the flat portion. It is preferable that some of the preliminary cambering is performed so as to be raised from the flat portion by deformation so as to return to a substantially flat state when stress of welding (particularly, laser welding) is added. The arcuate shape of the curvature from the flat state is less than 90 degrees, preferably between 10 and 80 degrees. A more preferred range is between 30 and 70 degrees. In some preferred embodiments, the metal sheet is stainless steel.

In another aspect, the invention provides a method of manufacturing a laminated device, comprising providing a subassembly, planarizing the subassembly, and welding the subassembly to a sheet or a second subassembly. Preferably, the subassembly is welded along the periphery except where there is an entry or exit opening. In some preferred embodiments, this method is combined with any of the methods described herein.

In another aspect, the present invention provides a method of performing a process in a channel layer that varies from tension to compression or vice versa, the method comprising: forming a first channel layer and a second channel layer directly adjacent to the first channel layer, Performing a unit operation in a first channel layer at a first time, the first channel layer having a first fluid of a first pressure and a second channel layer Wherein the first channel layer has a second fluid of a second pressure and the first pressure is greater than a second pressure, and performing a unit operation in the first channel layer at a second time, The second channel layer having a fourth fluid at a fourth pressure and the fourth pressure being greater than a third pressure.

At the time of tensioning, the pressure of the first channel layer is larger than that of the second channel layer, and the pressure of the first channel layer is smaller than that of the second channel layer. In the height direction, the layer is defined by a floor and a roof, and a tensile or compressive force is applied to the floor or loop of the layer. In some preferred embodiments, the first and third fluids are the same and the second and fourth fluids are the same, eg, the first and third fluids are Fischer-Tropsh ("FT") process streams process stream, and the second and fourth fluids are typically heat exchanging fluids that are water (or other heat exchange fluid) that is partially visually visible. Examples of other processes may include the synthesis of ethylene oxide, propylene oxide, methanol, ammonia, styrene, and hydro- genation and hydrotreating. Preferably, the method is performed in a laminate-type device having a first channel disposed in a first layer and a second channel disposed in an adjacent layer. More preferably, such a process is performed in a laminate-like device comprising a large amount (such as at least ten) of alternating layers of the first and second channels. In some preferred embodiments, the first channel comprises a press-fit insert, and in some preferred embodiments, the first channel comprises a catalyst preferably used for coupling with a press-fit insert. Surprisingly, it has been found that devices manufactured under such conditions without diffusion bonding or brazing (such as press-fit devices) can be well manipulated. Preferably, this method is defined when occurring during continuous operation, as opposed to occurring during termination or initiation. Such a process may occur, for example, when the solids are built up in the process channel and / or when the catalyst loses activity, and the process conditions are adjusted to accommodate changes in the reactor.

In another aspect, the present invention provides an apparatus for connecting a section of a device (as described below) with a halo. Accordingly, the present invention provides an apparatus having a first laminate-like assembly having a plurality of stack sheets, wherein the first side of the first laminate-like assembly comprises a plurality of inlets or outlets, 1 side and extends outwardly from the first side to cover the plurality of inlets and outlets. Typically, the enclosure is metal and welded on the first side of the first laminate-like assembly.

In some preferred embodiments, the apparatus comprises a second laminate-type assembly having a plurality of stack sheets, the first side of the second laminate-type assembly comprising a second plurality of inlets or outlets, Connects the inlet or outlet with a second plurality of inlets or outlets.

In another embodiment, the laminate-type microchannel device has a cross-sectional area defined by the sheet length X sheet width greater than 100 cm2 (in some embodiments, greater than 500 cm2) and from 0.05 to 20 cm over the section of the surface of the sheet in the microchannel device Having a length and a width, joined to the top plate to form a weld assembly having a linear density of laser welded joints of between about 0.1 and about 10 cm / cm < 2 > (preferably between about 0.1 and about 10 cm / (Preferably at least 90%, in some embodiments 100%) of the adjacent region of the substrate (e.g., having a major surface). In most cases, the sheet forms the surface of the subassembly within the larger apparatus. Two or more welded sheets are held together. In addition, welding may provide a seal between two adjacent internal flow channels. In this case, " 100% of the adjacent region " means the entire surface (not exactly 100% of the selected rectangular region), and similarly 50% and 90% means 50% and 90% of the entire surface. Preferably, the ratio of the length to the sheet width is greater than 2. Further, in addition to or in place of the linear density described, the apparatus of the present invention may have an internal weld (i.e., welded to the inside of the sheet, unlike the periphery) at least 10 times, preferably at least 100 times greater than the periphery ).

In another aspect, the invention provides a laminated microchannel assembly having a first sheet and a second sheet, wherein each sheet has a length and a width, the cross-sectional area defined by the sheet length X sheet width is greater than 100 cm < (In some embodiments greater than 500 cm2), the first and second sheets are substantially planar (the sheet may have some warpage but the sheet is not corrugated), the first sheet comprises an array of parallel microchannels, The microchannels are spaced from each other by a barrier wall (the microchannels may partially pass the thickness of the first sheet (e.g., the etch channel) or pass through the entire thickness of the first sheet) The second sheet is contiguous and includes welds (the welds are continuous or discontinuous) that follow the length of the barrier wall and join the first sheet to the second sheet.

The term " following " means that the weld proceeds in the same direction as the barrier wall and contacts the barrier wall.

In any of the methods described herein, an assembly may be formed by joining two or more weld subassemblies, and the method of welding two or more subassemblies to form an assembly may be the same or different welding methods . In some embodiments, the weld subassembly may comprise a weld formed by welding and other techniques.

In another aspect, the present invention provides a method of forming an assembly comprising welding an upper sheet and a lower sheet to form a plurality of channels disposed between the upper surface of the upper sheet and the bottom surface of the bottom sheet, And is used to form a seal between the channels in the plurality of channels. The top and bottom surfaces are the top and bottom surfaces of the laminate-like assembly. By way of example, the bottomsheet may comprise etched channels and the topsheet may be an unetched flat sheet. Preferably, the method of joining sheets to an assembly includes laser welding to seal between two adjacent internal flow channels. In any of the embodiments described herein, the channel is preferably a microchannel.

In another aspect, the present invention provides a repair process-treated welded substrate assembly for sealing a leak or hole in a primary welding step, wherein the repair process comprises the same welding methodology as the primary welding (typically laser welding) Or a secondary process such as TIG welding, pulsed laser, CMT or the like may be used to reduce the number of leak points in the welding substrate assembly.

In another aspect, the present invention provides a welded substrate assembly capable of holding a differential pressure greater than 100 psig at atmospheric pressure (more preferably greater than 500 psig and more preferably greater than 800 psig at atmospheric pressure). Welding assemblies are laminate-like devices in which the flow of fluid during operation is predominantly perpendicular to the sheet thickness. In the welded assembly, the seals to maintain differential pressure are not diffusion bonded or brazed.

In another aspect, the present invention provides a method comprising: obtaining a welded substrate assembly having a curvature greater than 1 cm when placed on a level table; and moving the welded substrate assembly upward to less than 1 cm above the flattened portion, A planarizing process; and a step of welding the planarized substrate assembly to a subassembly to form a laminate-type welding apparatus.

In another aspect, the present invention provides a fuel cell having a plurality of channels sealed by welding (the seal is not by polymer gasket, brazing, diffusion bonding or other prior art) and is pressurized with nitrogen at ambient temperature and 100 psig a leakage volume less than sccm nitrogen (preferably less than 1 sccm nitrogen) or a leak volume less than 0.5 psig per approximately 15 minutes.

In another aspect, the present invention provides a stack of sheets bonded together (preferably by welding), a continuous unfixed span constituting the voids in the stack of sheets, and a first non-adhesive span on opposite sides of the laminate- Further includes an array of reinforcing members that include two end plates and that maintain tight contact (e.g., via welding) with the major outer surface of the end plate and that extend continuously across the area aligned with the continuous un-mounted span Wherein the stack of sheets comprises a plurality of channels in the stack and at least one inlet and outlet connected to the plurality of channels. The array of reinforcing members is intended to resist bending in the stacking direction (i.e., in a direction perpendicular to the plane of the laminae). Preferably, the system further comprises a process stream containing hydrogen and / or hydrocarbons passing through the plurality of channels.

Further, the present invention can be provided for repairing a welded device. The apparatus can be retrofitted by removing one or more welds. By removing the plate at one end of the channel, the selected channel can be clogged to reduce high temperature spots or otherwise avoid channel defects. Alternatively, the device can be opened by removing peripheral welds, and then the subassembly can be removed or replaced. In some embodiments, the subassembly can be removed and refitted prior to reinsertion into the apparatus. After removal and / or replacement of the subassembly, welding may be used again to close the device. One or more surfaces of the subassembly may be coated with release layers such as grafoil or ceramic paper inserts between zirconia or yttria coatings or subassemblies to aid in unassembly.

The present invention also includes a method of holding a reactor by opening and removing or replacing a weld, and holding a catalyst, such as a particulate catalyst, a pin or corrugations, or subassemblies. Such a device can then be welded together side by side. The present invention also includes a welded structure for maintenance or repair.

In another aspect, the invention provides a laminate-like chemical processor comprising a plurality of sheets in a stack, wherein the stack has mutually vertical dimensions of height, width and length, the height is a laminate dimension for an open space in the stack, Is the longest dimension and the width is perpendicular to the length and the stack has at least one interface between the sheets in which the internal pressure is applied to the open space in the stack during operation and when N2 gas is fed into the void space of the interface through the inlet Wherein the open space has a width of at least 0.07 m and the outlet is closed such that the pressure increases at a rate of 30 to 50 kPa / min, the pressure of the void space is increased to 790 kPa, After the pressure is maintained, it is returned to the atmospheric pressure due to the release of the N2 gas, and thereafter a ratio of 300 to 400 kPa / min, sufficient to raise the pressure through the inlet Water is supplied and the outlet is closed so that the pressure increases and the pressure of the void space increases to approximately 3000 kPa and then the pressure is continuously increased at a ratio of approximately 100 kPa / min to 6000 kPa and then a pressure of 250 to 300 kPa The pressure is lowered to the atmospheric pressure, the pressure is lowered continuously, the water is drained to dry the processor, the N2 gas is supplied through the inlet so that the pressure increases at a ratio of 30 to 50 kPa / min at the interface, The pressure increases and the pressure of the void space increases to 790 kPa and the inlet is closed so that no more gas enters the void space and the device leaks less than 100 kPa for the following 15 minutes.

The present invention includes a protocol-tested processor as well as a processor with a claimed leak resistance. More preferably, the processor has a leakage resistance of less than 30 kPa for more than 15 minutes and, in some embodiments, a leakage resistance in the range of 1 to 50 kPa. In some embodiments, the open space has a width of at least 0.1 m, and in some embodiments at least 0.3 m. The processor may be of any shape, and in some embodiments the processor may be comprised of a stack of rectangular sheets, and in some other embodiments, the processor may be comprised of a stack of circular sheets.

Preferably, the processor does not have an end plate having a thickness greater than 3 cm, preferably no greater than 1 cm, and in some embodiments no greater than 0.5 cm. Preferably, the processor has a width of at least 0.3 m, and in some embodiments at least 0.5 m. The invention may comprise any combination of the features described through this description, for example the processor has a length and width of at least 0.3 m and does not have an end plate with a thickness greater than 3 cm. The interface may be planar, but is not necessarily planar.

In some embodiments, the apparatus includes internal linear welds and exoskeletons greater than 0.1 km, and in some embodiments greater than 1 km. Preferably, the laminate type chemical processor is held together by exoskeleton and welding. Preferably, the laminate type chemical processor is not diffusion bonded or brazed and has no gasket. Because of this leakage resistance (not due to the clamp) and because the clamp is not required to have a laminate-type chemical processor together, the processor need not be placed in a pressure-relief container. In some preferred embodiments, at least 60% by volume (in some embodiments at least 80% by volume) of the total laminate type chemical processor consists of microchannels and other void spaces.

In the leakage resistance test described above, "outlet closed" means that N2 is trapped in the void space except that it leaks through the interface between the sheets. It should also be noted that these tests apply to a single interface provided as an entrance or to an average sum of all interfaces. The apparatus is characterized in that at least one fluid conduit having a particular parameter meets the test if it meets the test and preferably the device has at least two fluid conduits which meet the test, . (For example, if the device has two inlets and two outlets each as a single outlet, both conduits meet the test). The device is under atmospheric pressure except for the area of the device connected to the inlet.

The invention also encompasses laminate-type chemical processors having one or more of the features (including any combination) described herein and having an exoskeleton. The continuous unattached span in the inner pressure boundary is the minimum distance between the predetermined attachment point between the laminae in the predetermined pressure exposed interface of the laminated device and the adjacent point of the attachment adjacent between the same laminae. In a preferred embodiment, the exoskeleton is welded to the device, and in another embodiment the exoskeleton is retained by brazing, gluing or other means.

The exoskeleton is superior to the clamp. The clamp can be easily removed. (The exoskeleton needs to be removed by cutting or grinding.) The exoskeleton welded reinforcing member also has a rectangular cross-section oriented toward the long side parallel to the load application direction to provide an increased reinforcement to resist bending stresses Lt; / RTI > This allows the use of a thin shell plate and reduces the load and cost of the material required to support the same rod. Clamps with thick plates with threaded fasteners are used at the exoskeleton location, but the plate needs to be strong enough against bending stresses because the threaded fasteners are not loaded in this direction. The threaded fasteners need to be sufficiently strong against the total tensile stress caused by the forces generated by the pressure acting on the plate. The exoskeleton provides additional support to the plate in both cases. Also, the clamps tend to loosen and break during repeated cycles.

The present invention also relates to a method of controlling a processor, comprising: passing a gas to a processor inlet to increase the pressure inside the processor to a first pressure; selectively monitoring the leak and selectively repairing the leak; Passing a fluid through the processor to increase the inner pressure to a second pressure; removing the fluid; passing a gas through the inlet of the processor to increase the pressure inside the processor to a third pressure; Measuring the leakage while the processor is maintained at the third pressure, wherein the second pressure is higher than the first pressure and the second pressure is higher than the third pressure. In some preferred embodiments, the fluid is a liquid. This method is superior to the single stage technique of loading the pressurized fluid into the device and testing for leakage.

In another aspect, the invention provides a laminate-type device comprising a stack of sheets joined by welding and a strain relief joint in a stack of sheets, The strain relief abutment comprises two adjacent sheets retained within the stack but is not substantially bonded together along the periphery of the two adjacent sheets.

Throughout this specification, " adjacent " means directly adjoining without interfering with the sheet.

The present invention includes any method of using any of the devices described herein, such as a chemical process using any device described herein as an example. As such, the present invention includes any apparatus for performing any of the methods described in this application. The invention further includes any combination of the methods and / or apparatus described herein. The sheet and (if present) inserts all preferably comprise a metal. In finishing equipment, the metal may be coated with a catalytic coating and / or a protective coating, such as a porous metal oxide layer having a catalytic metal dispersed on the metal oxide.

The present invention includes certain features and / or any broad technical idea described herein in various modifications and which can be reasonably deduced by one skilled in the art. By way of example, the apparatus of the present invention may comprise any combination of the features described herein.

The present invention is not limited to the specific technical ideas identified above, but includes any method, system, and apparatus described herein. The present invention may include any combination of any features and features described herein. The present invention also encompasses methods of chemical processes (including, for example, heat exchangers, chemical reactors, Fischer-Tropsch synthesis reactions), including, for example, Conditions), conversions, and the like. Although the process has been described with reference to a graph or a table, the present invention is intended to encompass within +/- 20% of the conditions, ranges and / or values described herein, more preferably about 10%, more preferably about 5% In the example, it includes a process with a value within about 1%. By way of example, the present invention provides for a CO conversion of between about 58 and 73% and a methane selectivity between about 8 and about 34% (the term " about " includes values within +/- 20% , The present invention includes a more limited method by the device structure and the present invention is otherwise limited to a system comprising both fluid composition and / or conditions and device characteristics And the system may be a device containing hydrogen gas and carbon monoxide at a temperature of 180 DEG C or higher.

Glossary of terms

An " assembly " is two or more plates joined together to form a laminate. The assembly may typically consist of a plurality of subassemblies and may be a precursor to a functional device or device. &Quot; Subassembly " is an " assembly " that is (or is intended to be) a component of a large laminate-like assembly. In some preferred embodiments, the assembly is completely sealed except at the inlet and outlet. The assembly need not be a fully functional device, i. E. It may be a precursor or intermediate product for a fully functional device. As an example, in some cases, the second trimming step needs to open the inlet and outlet flow passages. In some embodiments, the assembly comprises a plate having a width and length of a finishing device, and in some other embodiments the assembly may be cut into a plurality of subassemblies, or alternatively may be joined to form a large assembly. In some preferred embodiments, the assembly (or subassembly) has a thickness of less than or equal to 1 cm, in some preferred embodiments between 0.1 and 1.0 cm, and in some embodiments, of 0.2 and 0.4 cm. It is preferred that the plate in the assembly is substantially flat and the assembly has flat top and bottom surfaces.

Throughout this specification, the terms "plate", "sheet", "lamina" and "shim" are used interchangeably. The plate has a thickness of 1 cm or less, preferably 0.5 cm or less, more preferably 0.3 cm or less, and typically has a thickness of at least 0.02 cm.

&Quot; Exoskeleton " refers to an end plate that maintains intimate contact (e.g., via welding) with a major external surface of an end plate of a laminate-type chemical processor and interferes between the external pressure of the boundary of the array of reinforcement members and the internal pressure of the boundary Is an array of continuously extending reinforcing members. The array of reinforcement members is intended to resist bending in the stacking direction (e.g., in a direction perpendicular to the plane of the lamina). The exoskeleton is not a clamp and does not require screws or bolts.

The " gap " is the smallest dimension of the microchannel. Typically, in a lamina device, the gap is in the stack (e.g., height) direction. If the term " gap " is used, the preferred embodiment may be described in place of the height of the microchannel.

&Quot; Microchannel " refers to a channel having at least one internal dimension (wall-to-wall, catalyst is not counted) less than or equal to 10 mm, preferably less than or equal to 2 mm and greater than 1 탆 And 500 to 1500 microns in the 50 to 1500 micron embodiments are particularly preferred when used with the catalyst forming microparticles and preferably the microchannels are left in the dimensions of at least 1 cm, preferably at least 20 cm. In some embodiments, the length is in the range of 5 to 100 cm, and in some embodiments is in the range of 10 to 60 cm. The microchannel is also defined by the presence of at least one outlet and at least one other inlet. The microchannel is not a channel through zeolite or mesoporous material. The length of the microchannel corresponds to the direction of flow through the microchannel. The microchannel height and width are substantially perpendicular to the direction of flow through the channel. In the case of a laminate-type device in which the microchannels have two main surfaces (for example, a surface formed by laminated and bonded sheets), the height is the distance from the main surface to the major surface and the width is the height of the height. In some preferred embodiments, the microchannel is straight or substantially straight - which means that straight lines that are unimpeded can be drawn through the microchannel. (&Quot; unimpeded " means before inserting a solid catalyst, solvent, or other discrete solid material). Typically, the apparatus includes a plurality of microchannels sharing a common header and a common footer. Although some devices have a header and a single footer, a microchannel device may have a plurality of headers and a plurality of footers. Also, the microchannel is confined to the presence of at least one inlet that is different from the at least one outlet - the microchannel is not a channel through zeolite or mesoporous material. The height and / or width of the reaction microchannel is preferably about 2 mm or less, more preferably 1 mm or less. The side of the microchannel is defined by the reaction channel walls. These walls are preferably made of a hard material such as stainless steel or Fe-based superalloy such as Ni-, Co-, or FeCrAlY. The process layer may comprise a different material from the heat exchange channel, and in a preferred embodiment the process layer comprises copper, aluminum or other material having a thermal conductivity of greater than 30 W / m-K. The choice of material for the walls of the reaction channel may vary depending on the reaction the reactor is intended to be. In some embodiments, the reaction chamber walls are made of durable stainless steel or Inconel (R) with good thermal conductivity. Typically, the reaction channel walls are formed of a material that provides a primary structural support for the microchannel device. The microchannel device may be manufactured by known methods, and in some embodiments may be made by laminating interposed plates (known as " shims "), wherein the shims designed for the reaction channels are designed for heat exchange It is preferable that they are arranged mutually. Some microchannel devices include at least 10 layers (or at least 100 layers) laminated to the device, each layer comprising at least 10 channels (or at least 100 channels) . ≪ / RTI >

In some devices, the process channel comprises catalyst particles. Preferably, the particles have a size of 5 mm or less, in some embodiments 2 mm or less (longest dimension). The particle size can be measured by a sieve or microscope or other suitable technique. For relatively large particles, a sieve is used. The particulate material included in the process channel may be a catalyst, a solvent, or an inert material.

The heat exchange fluid may flow through a heat transfer channel (preferably a microchannel) adjacent to a process channel (preferably a reaction microchannel) and may be a gas or a liquid and may comprise a stream, a liquid metal, or other known heat exchange fluid The system can be optimized to have a phase change in the heat exchanger. In some preferred embodiments, the plurality of heat exchange layers are interleaved with a plurality of reaction microchannels. By way of example, at least ten heat exchangers are interleaved with at least ten reaction microchannels, and preferably have at least ten layers of heat exchange microchannel arrays interfaced with at least ten layers of reaction microchannels. In some preferred embodiments, if the process microchannel is an "n" layer, the heat exchange layer is an "n + 1" layer, whereby the heat exchange layer is laterally adjacent to all process layers. Each of these layers may comprise a simple, straight channel or a channel with a layer having a more complicated geometric shape. The present invention encompasses a system comprised of both devices and in which fluid is present in the device. In " bonding ", the heating process is used to bond the pieces to which the diffusion is produced to one of the elements from one piece so that the element diffuses near the interface (or near the one used as the interface before bonding) .

Brazing uses an intermediate layer interposed between the parts, and the intermediate layer has a lower melting point than the others.

Welding is used to bond or seal parts. Unlike brazing, welding does not require a low melting point, and the joints also need to be sealed to the periphery of the laminate-like device, although welding can be seen using welding wires of the same or similar material, Welded joints and some weld penetration depths are at different margins than through the product. The " welds " in the finishing piece can be identified by an expert - by way of example, the metallurgist can identify the weld by microscope inspection or other techniques known in the art.

&Quot; Joining " includes welding, joining, attaching, brazing. Joining is any process that joins two or more pieces together.

&Quot; Substrate assembly " is comprised of a plurality of sheets attached together to form a tacky laminate-like stack. &Quot; Substrate assembly " is sometimes referred to as a panel, and may comprise many sheets in a stack, which may consist of a top and bottom sheet forming a flow path, and more typically forming many flow paths.

The present invention also encompasses a method for performing unit operations within the apparatus described herein. &Quot; Unit operation " means chemical reaction, evaporation, compression, chemical separation, distillation, condensation, mixing, heating or cooling. &Quot; Unit operation " does not necessarily mean fluid delivery, although transfer occurs often, depending on the unit operation. In some preferred embodiments, unit manipulation is not just mixing.

The microchannel reactor has a length of less than or equal to 1.0 cm, preferably less than or equal to 2 mm (in some embodiments less than or equal to about 1 mm) and greater than 100 nm (preferably greater than 1 탆) (Wall-to-wall, without counting catalyst) of at least one reaction channel. The catalyst containing channels are reaction channels. More generally, the reaction channel is the channel from which the reaction occurs. The length of the reaction channel is typically long. Preferably, the length of the reaction channel is greater than 1 cm, in some embodiments greater than 50 cm, in some embodiments greater than 20 cm, and in some embodiments from 1 to 100 cm.

&Quot; Press-pit " describes the manner in which a fin (preferably a copper waveform) is disposed in a space within the apparatus. The press-fit pin is held in place by being compressed or seated within the device or within the subassembly. While there may be a small amount of tack welds, the press-fit pins are not brazed or welded in place at all contact points. Preferably, the press-fit pin is not held in place by any adhesive or any such chemical bonding.

Open space - refers to the space within the processor where there is no joined internal support that resists tension. The open space may include ribs or other structures providing a support in a compressed state, but such structures are not bonded to both sides of the interface and do not resist tension. The " open space " may exist as part of a larger space, but in the preferred embodiment the laminated device is welded or otherwise bonded only to the periphery.

The void space is a space within the apparatus that is accessible to the N2 gas that is passed through at least one inlet of the apparatus. The volume of the void space can be measured by evacuating the space for at least 10 seconds and measuring the amount of N2 gas passing the N2 gas into the space and filling the space.

Internal Linear Welding - Welding that joins two or more laminates together in the periphery of the outer peripheral welder.

In standard patent terms, " comprises " means " comprising " and these terms do not exclude the presence of additional or plural components. By way of example, it should be appreciated that where the device comprises a lamina, sheet, etc., the device of the present invention may comprise a plurality of lamina sheets or the like. In other embodiments, the term " comprising " may be replaced by the more restrictive term " essentially composed " or " composed ".

Brazing requires the addition of an interlayer material that melts at a temperature below the melting temperature of the constituent material. The intermediate layer becomes liquid during the diffusion brazing or brazing process. The liquid interlayer flows to fill the gaps or gaps so that the materials are bonded together. Further, since the intermediate layer flows and diffuses, the material from the intermediate layer can be diffused into the base material, and the material from the base material can be diffused into the intermediate layer.

As diffusion proceeds, the local composition of the intermediate material is changed. Further, when the temperature is lowered after the maximum temperature is reached, the liquid intermediate layer solidifies and fills the gap between the two base material layers. The solidification can proceed by temperature or composition. In the latter case, a melting point depressant such as phosphorous acid or boron may be added so that the intermediate layer melts at a lower temperature than the parent material. In a similar example, the diffusion joining device also forms a tight thermal contact between the heat transfer layers.

Figure 1 shows some components in which a process layer can be assembled.
Figure 2 shows a stack of components for forming sub-assemblies and stack subassemblies, each showing a process layer comprising a 3-pin insert.
Figure 3 shows a laminate stack in a compressed state, showing the corner of the device.
Figure 4a shows two plates forming a strain relief junction.
Figure 4b shows an expansion joint assembly welded to the top and bottom of the reactor core.
Figure 5 shows " halo " located outside the laminated device.
Figure 6 shows the cooling liquid surface and end plate chamfered surface aligned with the cooling liquid channel assembly.
FIG. 7 is a photograph showing the fillet welds (right in the figure) added to create a more uniform framework for the catalyst retention assembly, wherein the irregular pin shapes shown on the right side of FIG. 7 are generated by compression.
Figure 8 illustrates a process manifold for a reactor of the present invention, which shows a manifold (upper and lower) and a laminated reactor core (center).
Figure 9 shows a coolant manifold for an apparatus of the present invention.
Figure 10 shows the transition to the partial boiling and the stable performance of the pre-welded reactor.
11 shows a plot of the heat flow of the pre-welded reactor of the present invention at a contact time of 70 ms.
Fig. 12 shows pre-cambering into a pre-camber parallel to the coolant channel to reduce curvature.
Figure 13 is a photograph of a press fit pin adjacent to a ridge formed from a laser welding process in which the pin contacts the ridge and a crack in a small gap is observed between the fin and the heat transfer wall.
Figure 14 shows a laser weld line bonded to the top of the rib between parallel and adjacent coolant channels formed in the bottom plate.
Figure 15 shows a reactor with an external support (exoskeleton) and a laminate reactor core (left).
Figure 16 shows an assembly formed from four subassemblies joined by spot welding.
Figure 17 shows the apparatus of Example 8 with an external support, which is approximately 0.6-m X 0.6-mx 0.08-m.
18 shows a plot of the pressure cycle used in the hydrostatic testing of the process circuit of the apparatus of Example 8. Fig.
Figure 19 shows a plot of the pressure cycle used for hydrostatic testing of the coolant liquid conduits of the apparatus of Example 8;
Figure 20 shows a plot of the catalyst temperature along the centerline of the bed, with a pin height of 0.025 " (0.563 cm).
Figure 21 shows the catalyst temperature along the centerline of the packed bed of Example 10, with a pin height of 0.5 ".
Figure 22 shows the catalyst temperature along the centerline of the packed bed of Example 10, with a pin height of 1.0 " (2.5 cm).

The present invention described above provides a method for manufacturing an apparatus and an apparatus that can be manufactured by the method. The present invention further includes a method of performing unit operations in the apparatus. Unit operations may include chemical reactions, phase changes, mixing, heat transfer and sequestration. The device may be a microchannel or may be used in an apparatus having a large characteristic dimension. The dimensions of the characteristic microchannels are limited to a range of 10 mm or less, in the range of about 0.001 mm to 10 mm, preferably 0.01 mm to 2 mm, in some embodiments in the range of 0.1 to 2 mm.

In some embodiments, the method includes a first step of forming a subassembly from a top sheet and a bottom sheet, the bottom sheet being at least two sheets, wherein the bottom sheet may comprise an etch channel, And can be disposed between the top sheet and the bottom sheet. In some embodiments, subassemblies can be made from three or more sheets. There may be a small leak, but the fluid flowing through the first subassembly substantially remains within the first subassembly. In a second step, the first subassembly is laminated adjacent to the process layer, and the subassembly and process layer are pressed into thermal contact to form an assembly having at least two fluid passages.

One method of joining subassemblies produces short stacks of layers with two or more layers to create fluid passageways. In another embodiment, two or more layers may be joined to create an array of subassemblies or parallel fluid passages that allow fluid passageways for two or more fluids.

As one embodiment for forming the first subassembly, a shim or laminae top plate having a processed channel (the channel may be formed by etching) is bonded. The edges of the subassembly are larger than 95%, and preferably greater than 99%, and more preferably greater than 99.9% of the fluid entering from the first inlet, as opposed to leaking through unintended side or other paths. Sealed substantially along the edge to maintain continuity of the fluid passageway so that the amount is left in the subassembly from the first outlet and to prevent fluid from leaking out of the side. In other embodiments, it may have one or more inlets and / or outlets defined in a laminate geometry.

Preferably, the laminate is sealed along a top and / or bottom surface of a subassembly along a continuous metal or corridor of material, and in some embodiments at least 50% of the corridors are oriented at least in the longitudinal direction of the top and / And the seal typically follows a channel wall that separates the channel. When the subassembly is laminated, bonding occurs only in the region where the metal is in contact between the layers. It will be appreciated that the region having a fluid channel or a cavity for fluid traverse after the device is manufactured is not closed. Two materials are used for the contact portion for forming the sealing portion. Also, the joining of the subassemblies along one or more sides of the subassembly can be continuous along the fluid path or intermittent as required depending on the structural manipulation needs of the device. The fluid can be leaked or traversed from one parallel channel up to the first subassembly during quality control check or when tested with an operating device. A minor amount of traverse flow is less than 20% per channel flow, more preferably less than 10%, more preferably less than 2%, these percentages being based on a traverse flow average across all channels, or a traverse induction across selected channels can do.

The junction of subassemblies includes at least two layers, but may include three or more layers. In one embodiment, at least 20 layers are bonded to the subassembly. Methods for joining the first subassembly include but are not limited to laser welding, resistance welding, friction stir welding, ultrasonic welding, diffusion bonding, brazing or diffusion brazing or transition liquid brazing, adhesive bonding, reactive bonding, But is not limited to. In a preferred embodiment (especially with fiber lasers and Yb fiber lasers due to low energy input limiting the amount of metal warping after bonding), certain types of laser welding are used.

The method for joining the sides of the subassembly may be the same or different from the method for sealing the edges of the subassembly. In one embodiment, a Paser laser is used to seal along the periphery, and in other embodiments a pulsed laser is used. Other welding or bonding methods may be used to seal along the periphery (except in the area where the fluid passages enter or exit the layer).

The bonded or sealed subassembly is preferably checked for quality (" QC'd ") prior to lamination with the assembly. All subassemblies can be evaluated or the statistical sampling of the subassembly can be QC'd or the sampling of the subassembly can be evaluated for quality. The quality check may include a pressure test to check for leaks, a flow test to check for a pressure drop, or a dye test to check for a residence time distribution which may be an indication of flow between the intended sealed inner parallel channels .

Bonded or sealed subassemblies may be joined to an assembly by disposing or interleaving subassemblies bonded to a second array or second subassembly of fluid passages to create a device having two or more fluid passage sets.

Fluid passageways may include corrugated or pinned structures or other structures useful in chemical processes such as cellular structures such as foam, felt, wad, aerogels and honeycomb. In some preferred embodiments, the corrugation or fin structure creates a channel or chamber having an aspect ratio (height to width) greater than 1, the height being the distance between the two subassemblies, the width being a repetitive pin or wave shape The distance between adjacent legs (wavefronts) In yet another embodiment, the second fluid passageway may comprise any thermally conductive structure.

In a preferred embodiment, the second fluid passageway is a process channel, the first subassembly includes a heat transfer channel, and in other embodiments this function may be reversed.

An example of a component that can be used to construct a process fluid path is shown in FIG. 1, wherein a waveform is generated from a planar foil. The outside of the fluid passageway may be sealed with the edges of the strip (defined as peripheral strips or p-strips) or side bars and the use of support strips (s-strips).

The first subassembly is positioned or laminated between layers of the second fluid passageway (shown as corrugated layer). As shown in FIG. A single adjacent second fluid passage or a plurality of adjacent fluid passages (three shown in FIG. 2) may be laminated in each layer for the second fluid passage.

The process waveform may be bonded to the first subassembly using a thermally bonded adhesive or other material that uses welding along the apex of the pin or improves the thermal conductivity of the contact between the first subassembly and the second fluid passageway. In one embodiment, the two layers are bonded together during lamination and welding without the addition of any material to improve thermal conductivity. (In this embodiment, the absence of a brazed or welded joint is referred to as a "press-fit.") In another embodiment, an additive material is added to reduce the contact resistance between the first subassembly and the second fluid passageway. In another embodiment, thermal contact during a chemical process may be improved by use of a polyphase process, and a small gap or gap between the press pit process structure and the subassembly is filled with liquid during processing through the capillary force. The liquid may preferentially fill the gap and enhance the conductivity of the composite structure when it is acting as a chemical process unit.

After laminating the hybrid stack including both the first subassembly and the second fluid passageway, the device of the present invention is engaged to form a stack by methods such as external welding, adhesive, and reactive bonding. Stack welding can use different types of welding methods including TIG, MIG, laser welding, electron beam welding, and the like. External welding is primarily automated for reproducibility and cost savings. Solder may also be an option to bond the perimeter if the temperature and pressure of the chemical processor is sufficiently soft to allow conduction to the solder.

Prior to bonding the final assembly, the stack can be compressed to reduce the pores by contacting the layers, resulting in a final device bond. Compression may occur, for example, through the use of a clamp-type fastener applying a load to the bolt assembly or through the use of external pressure to apply a load to the stack. The press-fit waveform can be deformed during compression and can remain deformed after compression is removed.

The subassembly requires planarization prior to lamination. One method of planarization includes roll leveling of laser-welded subassemblies using a leveling machine. This method reduces deformation when a 6 "x 24" (15 cm x 60 cm) panel is used. These panels have one dimensional change along the length of the weld line. Roll planarization is unsuccessful in a 24 "x 24" (60 cm x 60 cm) subassembly where part is deformed in two directions (bowl or three-dimensional parabolic shape). Conventional leveling machines are used to planarize twisted portions, but as a result, laser welds are broken. A soft hand roller was used instead of rigidly bending the twisted portion in a more flattened and less aggressive manner. A non-rigid hand roller somewhat reduces deformation but does not reduce it to a flattened condition defined as being less than 1 cm at any corner when placed on a flat surface. Thus, a smooth planarization can produce a superior device especially for assemblies having different widths and lengths (i.e. sub-assemblies that are not square). In some preferred embodiments, planarization is performed on subassemblies having a width of less than about 15 cm, and in some embodiments, a width of 10 to 20 cm.

The V-grooves are advantageous between sub-assemblies because the weld fillets can be applied to fill the V-grooves. The subassembly may protrude slightly from the sidebar or edge strip region. In another embodiment, the subassembly is substantially flush with the edge strip. In fact, the same height means an increment of sub-assembly thickness of less than five. By way of example, if the subassembly is 0.01 "(0.025 cm) thick, the edges of the subassembly do not protrude or engage more than 0.05" (0.125 cm) from the edge of the edge strip. For 0.06 "(0.15 cm) subassemblies, the offset from the same height edge is not greater than about 0.3" (0.75 cm), and the degree of good offset is no greater than 0.06 "(0.15 cm) from the flat portion, 0.06 "(0.15 cm) overhang or 0.06" (0.15 cm) overhang as a preferred embodiment.

A key advantage of the hybrid method of this manufacturing is to reduce the need for diffusion bonding and / or brazing during surface treatment. The surface must be clean and flat, and have tight tolerances that are tightly fitted during quality diffusion bonding and / or brazing. The elimination of the brazing and / or bonding step eliminates the need to make the large device hot as required for diffusion bonding and / or brazing. The energy required to heat and cool a large apparatus is important when it is required to heat and cool a large apparatus to a junction or brazing time without causing excessive mechanical thermal strain and thereby deforming. For devices made from stainless steel consisting mainly of planar internal laminas, the internal thermal gradient from the outermost to the center point should be less than about 30 ° C and greater than 500 ° C to prevent mechanical deformation of the layer. For devices with cross-sectional areas greater than 0.5 m x 0.5 m, it may take several days to heat the device and several days to heat the device when brazed or bonded in a vacuum-based thermal process. This required processing time and surface preparation of the part increases the overall cost of the reactor.

The method for manufacturing a device of the present invention eliminates the need for a diffusion bonding and / or brazing step of the reactor. As a result of the process of the present invention, reactors are manufactured to a high quality while reducing costs and time.

The surprising result from the press-pit apparatus of the present invention is the effect of the contact resistance between the layers. The press pits of the layers do not guarantee intimate thermal contact, but become worse as the size of the device increases and the planarization initiation portion is not perfect. Heat is removed between the first subassembly and the second fluid passage through the low quality contact area separating the fluid stream. In a diffusion bonding or brazing device, each layer makes close thermal contact by the nature of bonding and / or brazing, while the roughness and / or partial irregularity or additional deformation of the localized surface reduces the heat transfer effect between the layers. The importance of thermal contact between the layers may vary depending on the process operation requirements of the reactor or device. In some embodiments, the internal void between the two layers is filled by the process fluid during operation. In another embodiment, a thermally interconnection material such as a deformable solid such as graphite or a comparable interlayer or a liquid or putty or adhesive reduces the contact resistance between the two fluid layers (at least one being a press pit layer) To the press-pit layer.

In some embodiments, no intervening thermal contact layer is required. The Fischer-Tropsch reaction is tested in the reactors of the present invention without the use of an interference layer to improve thermal contact between the coolant side of the laser welding subassembly and the process side wave form. This practice is in fact matched to what is measured from all brazed reactors of virtually similar design.

Also, reaction or unit operations involving any of liquids and / or hydrogels, including, but not limited to, hydrocyanation, hydrocracking, or hydrotreating reactions, may interfere with the thermally conductive layer between the first subassembly and the second fluid path It is not necessary to do this. Such fluids have good thermal conductivity and good thermal conductivity can be obtained when such fluids fill the voids. The liquid has sufficient capillary pulling force to be transferred to the pores between the fin and the adjacent heat transfer surfaces. In addition, the surface tension of the oil is lowered to copper than stainless steel, which further strengthens the capillary pulling force of the oil or wax in the case of the Fischer-Tropsch reactor, and is lowered to the pore between copper and stainless steel (or other metal) during the reaction. The high surface tension of the oil or liquid wax for stainless steel may not have the same effect or may require a substantially small gap to deliver the liquid to the air. Copper pins are also more tolerant of irregularities.

It can be seen that the reactions involved in the oxidation reaction require interfering with the thermally conductive material. In one embodiment, the washcoat catalyst or other holding fluid of the second fluid path can also function to interfere with the thermally conductive material when filling the gap between the two layers to assist in the heat transfer and thermal control of the reaction. have.

Strain relief joint

A strain relief joint made up of two plates welded together can be added to the stack to reduce the strain imposed on the weld joining adjacent layers in the core. The joint is designed to open (expand) when the reactor is compressed during operation. By doing so, the sealing welds located on the reactor exterior surface remain unmodified, increasing the life of the unit.

The expanding joint typically consists of two metal plates with the same width and length. By way of example, the plates are ~ 24 "(60 cm) wide by 24" (60 cm) long to match the dimensions of the other plates of the stack, as described in this description. Preferably, the bottom plate is thinner than the top plate at the strain relief junction and the base plate (i. E., The plate is coplanar with and in contact with the stack) is ~ 0.25 " (0.625 cm) thick and the top plate Plate) is about 0.04 " (0.1 cm) thick. In some preferred embodiments, the base plate includes a plurality of holes, and the top plate is monotonous. The plates are positioned on top of each other and aligned at the edges. The plate is then welded through a laser welding process. The edges of the expansion joint plate are not continuously welded, which allows movement of the plate during operation of the apparatus. Preferably, the weld pattern is such that edges of the plate are not joined, except at the corners. This allows the edges to separate and separate the different internal streams from each other during operation if necessary to achieve any expansion of the reactor without deforming the sealing welds to the reactor. The holes in the base plate allow the welds to be individually checked for leaks prior to assembly with the reactor. If the expansion joint passes a qualification test, the base plate may be filled using a standard TIG welding process.

Preferably, the two expansion joint assemblies are used in a complete reactor core, one at the top of the core and one at the bottom. Preferably, one plate of the strain relief joint is positioned relative to the reactor core and welded to the coolant subassembly around all of the periphery.

During operation, the reactor is compressed. Compression is the result of some elastic stretching of the outer support. If no strain relief joint is present, it will cause a corresponding stretch of the reactor core itself and, as a result, stress is induced in the strain generated in the weld. By the presence of the expanding joint, the elastic stretching proceeds to the opening of the joint to relieve the strain on the weld.

The weld selection step on the halo is made from two or more parts or adjacent hollow squares or rectangular metal rings welded together to form an adjacent ring projecting over the face of the core. The halo creates a structure in the middle of the final operating manifold (macro manifold) and the device, so that during the refurbishment, the connection between the macro manifold and the device can be cut off and re-welded or bonded during continuous operation. The use of such halo has particular advantages as a means for removing or refining the catalyst contained within the reactor core. Halo, a device with halo, a method for producing an apparatus with halo and a method for using an apparatus with halo are additional aspects of the present invention.

Yes

Example 1: Welding Assembly - Welding Reactor - Press-fit Conductive Pin

Welding reactors are fabricated and operated to demonstrate performance equivalent to a braze reactor using Fischer-Tropsch as a test reaction. The reactor is operated on the stream for over 2000 hours and the press-fit contact of the catalyst containing process fin to the subassembly appears to be sufficient for reactor performance and is matched to the performance from the same design of the braze reactor.

Device Description

A two-layer Fischer-Tropsch pre-welding apparatus is designed and manufactured to demonstrate the manufacturing process of the present invention. The multi-channel microreactor design consists of two friction repeat units interposed between the three coolant repeat units. The coolant channel has a cross flow direction for the process channel.

The process channel is produced with a copper waveform of 3.75 cm (3 inches) wide and 0.635 cm (0.256 inches) and 15.75 cm (6.2 inches). The thickness of the waveform is 0.015 cm (0.006 inches). The final device has 274 processes in two layers. Each of these channels has an average dimension of 0.0372 inches wide, 0.25 inches high, and 3 inches long. The pins are as large as 0.006 " (0.015 cm) from the normal dimensions of the adjacent edge strips to ensure good thermal contact.

The coolant channel of the device is a Yb fiber laser (IPG Model YLR-600-SM: 600-watt) with a weld thickness between 50 and 150 microns through a 500 micron top plate that passes undamaged through a 1000 micron bottom channel plate A laser welding subassembly bonded to the top plate with a laser, a terrestrial fiber laser, a 1.07 micron wavelength). The subassembly is produced from two shims comprising an upper or cover sheet or wall that is accompanied by a channel shim having flow channels for the heat transfer fluid. These coolant subassemblies are laminated to a ~ 3 "L x 10" W x 2.7 "H (8 cm x 25 cm x 7 cm) device and sealed to the periphery by fusion and fillet welding.

After welding of the core component, the apparatus was equipped with 66.5 grams of a highly active cobalt catalyst from Oxford Catalysts, Limied, 120 grit material supplied by an Atlantic Equipment Engineer Lt; RTI ID = 0.0 > silicon carbide < / RTI >

The final fabrication step consists of header and footer welding and support welding (macro scale, ie for cooling liquid and process channels for external connection to large piping). Support welds have a structural need for designs that allow the device to be operated safely without the need for a pressure suppression system (PCS).

Core Component Manufacturing

The pre-welded reactor comprises two process layers with a copper corrugation and two stainless steel edge strips. The three coolant layers are fabricated as subassemblies that are interleaved with the process layer and are laser welded together with a stainless steel top plate or wall and a stainless steel channel shim. Wall shims are cut from 0.020 "(0.05 cm) thick stock SS sheet material The coolant shims are 0.015" (0.0375 cm) deep x 0.017 "(0.0425 cm) wide for non-straight section and 0.040" Non-straight and straight lines formed through partial PCM (photo-chemical machining) of 0.0040 "(0.1 cm) thick SS sheets to produce 0.020" (0.05 cm) depth x 0.100 " The non-straight section is made in a wavy or serpentine pattern with 22 turns Laser welding is done on each rib and the overall length of the shim is between the outermost channel and the periphery to seal the channel from the outside, The subassembly is then checked for leaks. The performance of the leak check subassembly allowed for identification and repair of leaks prior to assembly of the device is avoided and the assembly of the damaged device is avoided. As with the subassembly, preassembly Another benefit of having a layer of coolant liquid is the reduction of the portion for stacking into the assembly by at least about 20%.

Edge strips to the process side are fabricated from standard material thicknesses (normally 0.250 " thick) (0.625 cm) and require minimal machining to cut only the length and width and chamfer the edges. , And the thin coil is regularly bent to produce a repeating pin structure. The fin is 0.256 " (0.64 cm) high. The end plates require minimal machining to achieve good length, width and chamfer. A stock material may be used since a tight thickness tolerance coupled to the brazing device is not required in all parts.

The device assembly of Example 1

The core of the device (i.e., the process and the coolant layer interposed between the two end plates) is created by laminating the components to produce an interposed process and a coolant layer. The number of process layers is determined by the good capacity of the FT product, but the number of coolant layers is at least one more than the number of process layers so that the process layer has a coolant layer on both sides. During the lamination process, the fasteners are required to align the parts and to maintain this alignment throughout the assembly and in the initial welding step. The clamp fixture is designed to create a platform in the stack and to secure the transfer stack core to the welding stage. The clamp fixture is composed of two plates of a long plus sine shape. Each has four slots for fixing all the screw rods of 1/2 "(1.25 cm). The support plate is placed under the floor clamp to create a room for the nut on the bottom side and all threaded end For alignment of the process surface, four straight edges are held in place on either side of the clamp fixture with the c-clamp. The fifth straight edge is used to align to one of the two cooling liquid surfaces.

Due to the stacking and alignment fixture setup, the first step is to position the end plate on the clamping fixture and center it between the process surface straight edges. The coolant surface straight edge is then fitted in place before the first layer is laminated. The stacked first layer is a cooling fluid assembly. The process layer is disposed between the two coolant layers. The subassembly is lowered onto the end plate and is slid to a position against the straight edge of the cooling liquid surface and centered between the raised surfaces on the end plate (Figure 6). Once the alignment is satisfied, Respectively. At this stage, the edge strip is flush with the cooling liquid surface. For fusion welding, this seals the interface between the edge stream and the coolant subassembly and between the edge strip and the end plate. The first edge strip is disposed at the same height as the coolant straight edge following the location of the corrugations and then positioned flush with the second edge strip. The corrugations are positioned closely to the coolant first edge strip and centered on the coolant subassembly. The second edge strip is positioned closely to the waveform and the alignment with the other cooling liquid surface is checked. Alignment is acceptable if the edge strips on both coolant surfaces are +/- 0.010 "(0.025 cm) of the adjacent layer. The process of laminating this coolant subassembly and process layer is repeated one or more times and then on another coolant assembly The last end core plate is located on the upper end plate and has the same height as all four process side linear edges and the cooling liquid surface straight edge, .

There are two welding steps to complete the core welding. The first welding step is to form fillet welds along the 2 " (5 cm) wide edge strips. There are six V-grooves in each process face filled with fillet welding. The initial fillet weld is designed to stop the short of the end of the edge strip closest to the corrugation to avoid damage. The fillet weld is then applied to the edge of the edge to match the catalyst retention assembly, The front and rear views of this fillet weld are shown in Figure 7. Upon completion of this fillet weld, the clamping fixture is removed to allow it to approach the cooling fluid surface for the next welding step. There are three coolant subassemblies on the liquid surface, six seams (one core above and below each assembly) ) Each have a fused weld at the full length of the face. The core is then awaited for cleaning and preparation for catalyst loading.

Cleaning and Catalyst Loading

Prior to catalyst loading, the process side of the apparatus is cleaned and the catalyst retention assembly is welded in place.

Prior to catalyst loading, the catalyst retention assembly is inserted and welded in place on one process surface to retain the catalyst, and the core is oriented in a vertical position. The catalyst retention assembly consists of four parts, a screen, a screen retaining ring, a foam and a foam retaining ring. The screen serves to retain the catalyst in the apparatus. The screen retention ring is a thin SS frame that holds the screen tightly against the process surface in situ. Small welding around the periphery of the screen retention ring secures the screen in place and ensures good catalyst retention in the apparatus. Fillet welding on edge strips is also applied to the inner edge of the edge strip to provide better catalyst retention. An aluminum shielding plate may be used to protect the copper waveform during peripheral welding of the screen retaining ring.

Catalyst loading is performed by having the catalyst retention assembly in place in one process plane. Loading is a four-step process. The four end channels (one at each end in each process layer) are considered as inactive channels because they have partial pins and are blocked by the screen retaining ring. This channel is completely filled with an inert material, silicon carbide (SiC), of approximately the same particle size as the catalyst in the first loading step. The remaining channels are loaded with ~ 0.665 "(1.6625 cm) SiC, ~ 1.5" FT (3.75 cm) catalyst bed at process inlet and ~ 0.75 "(1.875 cm) SiC at process outlet. Depth of each of the three layers The material is loaded into the apparatus in small increments by laying the side of the apparatus (end plate) with a rubber mallet to densify the loaded material. Each incremental rod and after densification , The gauge height pin is used to measure the depth of all channels. When such a process is completed for any given layer, the material should be as close to the PABD as possible (the packed average bed density The loading is completed when the top layer is left at the same height as the edge strip after densification. The catalyst is removed by ultrasound, As all three layers (two layers of SiC, one layer of catalyst) are loaded, the catalyst retention assembly is installed and the same procedure as described above for the other process surfaces is used And welded to the open process surface of the apparatus.

Final Welding for Example 1

The three final welding steps complete the installation of the process manifold, the coolant manifold and the support plate. Unlike the brazing device, only the internal components for the totally sealed front welded FT device are the coolant subassemblies. The remaining components (corrugations, edge strips and end plates) are all adhered to adjoining components at the periphery. The support plate provides the necessary structural support for the device to maintain integrity under large differential pressure during operation. In addition, the backing plate serves to replace the pressure restraining shell (PCS) used in the brazing device. Two outlets are used for the coolant to remove steam and liquid water separately.

The process manifold is made of stainless steel 304L and is approximately 9.1 "(2275 cm) long by 2.7" (6.75 cm) wide by 1.9 "(4.75 cm) deep. Approximately 8.1" x 1.7 "x 1.2" 4.25 cm x 3.00 cm) encapsulates the process waveform opening as a whole and provides a support for the catalyst retention mechanism. The manifold is welded around the perimeter of the core using a conventional TIG welding process. A one inch diameter tube extends from the center of both process manifolds to allow process gas to enter and exit the core. The manifold is designed with a minimum wall thickness of approximately 0.5 "(1.25 cm) to support the process during operation (see FIG. 8).

The coolant inlet manifold is made of stainless steel 304L and is approximately 5.6 "(14 cm) long by 2.7" (6.75 cm) wide by 1.7 "(4.25 cm) deep. Approximately 4.8" x 1.9 "x 1.3" x 4.75 cm x 3.25 cm) is designed to encapsulate the coolant inlet channel as a whole and uniformly distribute the coolant over the coolant inlet surface. The manifold is welded around the perimeter using a conventional TIG welding process. One inch diameter tube extends from the manifold to allow the coolant to enter the core. The manifold is designed with a minimum wall thickness of approximately 0.38 "(0.95 cm) to support the coolant pressure during operation.

The coolant outlet manifold is made of stainless steel 304L and is approximately 5.6 "(14 cm) long by 2.7" (6.75 cm) wide by 4.4 "(11 cm) deep. Approximately 4.8" x 1.9 "x 4.0" x 4.75 cm x 10 cm) is designed to encapsulate the coolant outlet channel as a whole and to allow the coolant to exit the reactor core without interference. The manifold is welded around the perimeter using a conventional TIG welding process. Two 1 " (2.5 cm) diameter tubes extend from the manifolds from opposite sides. The upper tube allows steam to escape, and the bottom tube allows liquid water to escape. Lt; RTI ID = 0.0 > 0.38 " (0.95 cm). ≪ / RTI >

The two coolant manifolds are directed to and welded to the solid metal core end plate and the process manifold. By doing so, the core weld is encapsulated entirely within the parameters of the process and coolant manifolds and is thus not directly exposed to the outer surface of the reactor (see Figure 9)

The support (exoskeleton) is then added using a conventional TIG welding process. There are four sets of supports surrounding the reactor in the vertical (process flow) direction and one additional set extending horizontally over the coolant outlet manifold. Each set of vertical supports consists of two identical pieces of stainless steel 304L having an approximate overall dimension of 8.8 " (22 cm) length x 3.3 " (8.25 cm) height x 0.25 " (0.625 cm) Each set of supports is welded together at a point where the three ends are in contact and sewn to the reactor around the periphery. The four sets of supports are spaced from each other by about 2 " (5 cm) (6.75 cm) from the edge of the manifold, while four vertical sets of supports are provided to support the reactor core and the process manifold, and the horizontal set is provided with additional support for the large coolant outlet manifold The two horizontal supports are made of stainless steel 304L and are approximately 5.4 "(13.5 cm) long x 2 '(60 cm) high x 0.25" (0.625 cm) thick. They are either one of the coolant outlet manifolds Center on the side And, it is welded to the end plates, and manifolds of the outermost vertical support and the core.

Experimental Setup for Example 1

Process side

The flow and composition of the syngas fed to the Fischer-Tropsch composite microchannel fixed to the bed reactor is controlled by setting the flow rate of the individual gases (carbon monoxide, hydrogen and nitrogen) using a Brook flow controller . The gas is supplied via activated carbon and molecular sieve 13X traps to remove any impurities. This feed is preheated in a stainless steel microchannel heat exchanger before entering the reactor. The reactor is enclosed in a calm-shell 3000 W Watlow heater to isolate it to minimize heat loss. The manipulation data is measured using a pressure transducer and a 316SS sheath k-type thermocouple.

The product stream is sent through three collection vessels at elevated pressures and is cooled in the stage to provide a mildly substantial isolation from the heavy hydrocarbon product in liquid phase. The first product tank (maintained at ~ 100 ° C) and the second tank (held at ambient temperature ~ 25 ° C) collect most of the product. Liquid and heavy hydrocarbon (wax) products are collected in the first tank, liquid and clean liquid hydrocarbons are collected in the second tank. The waste gas from the third tank is evacuated.

The product gas sample was collected via a sample port located immediately downstream from the reactor and immediately upstream from the first product collection tank and was loaded with Agilent M200H mannitol gas having two columnar bodies 5A and Plot Q ≪ / RTI >

Coolant side

A 20gal (76L) carbon steel tank is used to store the coolant. Vertical chemistry is maintained with the addition of OS5300 and Optisperse AP302. The tank is pressurized with nitrogen to maintain the steam loop pressure. Cat Pump (Model 231.3000) is used to pump water through the cooling circuit. The Appleton FLSC-62A flowmeter is used to measure and control coolant flow. Feed water passes through a 25 cm, 5 micron particulate filter bank and a 60 micron Swagelok filter before entering the reactor. The reactor steam outlet is connected to a nitrogen source to control the coolant pressure, and the water (separated from the coolant foot) flows into the 2 liter stainless steel and gelock vessel used to maintain the appropriate level in the system.

Performance data

The pre-welded reactors of the present invention are operated for ~ 2150 hours without a pressure-relief shell. The reactor has an externally welded exoskeleton to provide a pressure support for this high pressure reaction.

The reactor is operated in increasingly stringent operating conditions at a point of thermal runaway (about 70 millisecond contact time on the process side). After thermal runaway, the reactor is regenerated to measure the extent of damage to the catalyst. After regeneration, the catalyst is restored to nearly 50% of its initial activity.

Start-up and demonstration testing

The start-up of the reactor is as follows. After the catalyst activation is complete, the reactor is cooled to ambient temperature and then pressurized to 350 psig (2413 kPa). The coolant is introduced into the coolant loop at the target flow rate and the reactor is heated to a synthesis gas inlet temperature of ~ 170 ° C. At a later stage, syngas flow is initiated and the reactor is heated to the target operating temperature.

At the end of the start-up, the reactor is a _{H 2: CO = 2.0, P} = 350psig (2413 kPa), ~ 16.8% dilution, to obtain the conditions of CT ~ 290 ms. Two rows of thermocouples are welded to the outer reactor surface at approximately 1.17 cm (0.46 in) and 3.2 cm (1.26 in) from the catalyst bed starting point (3.20 cm and 5.23 cm from the reactor inlet). The full welded microchannel reactor of the present invention does not cause complete isothermal reactor operation, but the measured temperature gradient is less than about 5 ° C. Also, the internal gradient for the FT catalyst is expected to be larger than the measured thermal gradient measured at the reactor wall without being measured.

The temperature profile across the reactor surface is controlled within 2 ° C of the average temperature.

The performance of the inventive (pre-welded) reactor and a direct comparison with the braze reactor are tabulated in Table 1 below based on the same Fischer-Tropsch catalyst.

Table 1: Comparison of the (pre-welded) reactor and the braze reactor of the present invention

* Notice: The temperature measurement position is slightly different between the two reactors. In the braze reactor, the temperature was measured at the surface of the cooling liquid shim, while in the present invention (pre-welded) reactor, the temperature was measured outside the reactor wall surface.

The time for the stream performance of the pre-welded reactors of the present invention is comparable to other brazes and single-channel reactors under the same conditions.

The product wax collected during this demonstration period is analyzed for carbon number dispersion. The results show excellent agreement from initial test to wax under conditions similar to short and long single-channel reactor and braze pilot-scale reactor tests.

Robust to process upsets

In a stream of approximately 211 hours, breakage of the coolant flow meter leads to an interlock situation. Coolant flow meter damage generates a zero flow warning that causes a back-up pump (despite normal pump function), resulting in significantly higher cooling fluid flow to interlock. The system is reset within 5 minutes. The reactor is cooled to ~ 197 ° C during this period. The CO and H ₂ flows are immediately turned on when the system is reset. After 2 minutes, the N ₂ flow is turned ON. Reactor temperature begins to rise immediately (within 9 minutes of system reset) (the coolant channel is not draining, so the pool of water from the previous operation is on the coolant side where it can begin to evaporate) The temperature reaches ~ 240 ° C with no coolant flow. (9 minutes from system reset), the coolant pump is started manually. The reactor temperature begins to drop to a normal level. (37 minutes from system reset) Within 28 minutes, this situation is under control and the reactor is cooled to ~ 192 ° C. Thereafter, the reactor temperature gradually increases to a value before the interlock (~ 206.6 ° C). During reset, the H2 flow is set to a value higher than the target, resulting in a ratio of H2: CO (instead of 2.00) to 2.17.

These experimental results surprisingly indicate that there is no thermal runaway in the few seconds of loss of the supplied coolant, and 9 minutes between the loss of coolant and the reactor temperature rising above 40 ° C. The catalyst performance is retracted to the expected level after achieving the target temperature after the coolant has been resumed. The high ratio of metal reactor block volume to catalyst volume creates a heat sink to absorb the heat of reaction for several minutes while the system upset is reversed. This is particularly advantageous over conventional tubular fixed bed FT reactors, whereby the temporary loss of cooling fluid results in loss of catalytic performance and thermal runaway. The FT reactor of the present invention produces a suitable transient buffer for the undesirable column upsets as shown to return to expected performance after 9 minutes without cooling fluid flow. The reactor temperature increases as expected, but the high thermal capacity of the metal structure causes the catalyst to remain permanently sintered.

For the entire weld pilot reactor, the catalyst volume is ~ 7% of the total volume [63.1 ml catalyst in a 0.934 L reactor block - 10 "x 3" x 1.9 "(25 cm x 7.5 cm x 4.75 cm) : For the reactor volume for a catalyst volume of 1, a heat sink time of 9 minutes without cooling liquid appears acceptable.

In large devices where the ratio of reactor volume to catalyst volume is less than 14: 1, more typically less than 10: 1, more preferably less than 3: 1 or 2: 1, the acceptable time without cooling fluid flow is 9 minutes Less than 5 minutes, and in some embodiments less than 30 seconds. In some preferred embodiments, the ratio of the catalyst volume to the reactor volume is between 2 and 60%, in some embodiments the reactors are between 5% and 40%, and the total reactor volume is the channel, channel wall, Folds, and outer walls, but does not include an outer pipe or pressure-relief container.

Also, in the ~ 346 hour stream, the carbon monoxide flow controller is changed due to the low dry test meter (DTM) which is the readout for the outlet flow. The new CO MFC is set to a lower value than the target, and the H2: CO ratio increases. During a continuous period of 17 hours (from stream to 363 hours), the H2: CO ratio is ~ 2.36 and the CO conversion increases to a value greater than 85%.

Surprisingly, it has been found that the ratio of deactivation does not increase even at high CO conversions, and performance is restored to the previous level after the feed rate adjustment. The reactor has a robust operation with a range of conditions including high levels of H2 to CO. Conventional tubular fixed bed FT reactors do not respond well to rapid changes in heat output, including a sudden increase in the heat released from the reaction. The reactors of the present invention continue to operate in a stable manner, such as a reaction heat that is increased by the ratio of H2 to high CO. After the synthesis gas ratio has recovered to the target dimensions, the performance returns to the expected performance.

Partial-boiling judgment

In the next phase of the reactor certification of the present invention, the partial boiling of the cooling liquid is tested and is followed by a linear seam between the parallel wavy characteristics, as opposed to conforming to the wavy profile, 22 wave turn) thermal control and stability of the entire welder reactor of the present invention with only sealed boiling flow control characteristics is demonstrated. There is one wave characteristic section per cooling fluid flow channel and a simple linear seal between the parallel cooling fluid channels is sufficient to maintain stable operation in the high temperature flux partial boiling controlled reaction and by way of example, But does not bypass the following turns, thereby lowering the front pressure drop and potentially causing unstable cooling fluid flow during boiling.

Operating conditions (H ₂ : CO = 2.0, P = 350 psig (2413 kPa), ~ 16.8% dilution, CT ~ 290 ms) corresponding to the "home" condition are maintained. Stream at ~ 634 hours, and the coolant flow is lowered from ~ 2 LPM to achieve an output steam quality of ~ 1-3% boiling. The reactor temperature is adjusted to maintain ~ 70% CO conversion. At ~ 679 hours in the stream, the coolant flow rate is reduced to 0.4 LPM and an exit steam quality of 1.5% is achieved. The reactor temperature is lowered to 204.3 캜 to maintain the target CO conversion. Performance is in fact similar to single phase operation. In addition, the use of partial boiling allows the reactor to be operated with high net heat flux or heat generation from the FT reactor operated at contact times lower than the initial 290 ms.

Partial boiling conditions are maintained in the stream for ~ 300 hours, during which performance is summarized in FIG.

The reactors of the present invention are operated stably for more than 250 hours on the stream with a partial boiling of water with an outlet steam quality of 1.5% at " groove " conditions of 290 ms to substantially match the performance of a single phase cooling fluid.

Thermal stability determination

In subsequent test steps, the reactors of the present invention can effectively remove heat by increasing the heat duty of the reaction (which treats many syngas by reducing the contact time while maintaining the CO conversion by adjusting the temperature) Is tested. It should be noted that the process pins are only in press-fit contact with the coolant subassembly. The contact resistance of the press fit process pin with the coolant subassembly does not substantially change the performance of the heat dissipating FT reactor. In addition, the press pits of the pins to one wall may have a height of from 0.5 to 5 mils (0.013 to 0.13 mm) and a width of from 0.025 to 0.508 mm (1 to 20 mils), resulting from the laser welding method of manufacturing the sub- It is hampered by the raised rib or nub.

The process reaction contact time is stepped down from 290 ms to ~ 70 ms, while maintaining the operating pressure of 16.5% dilution, H2: CO = 2, 350 psig (2413 kPa). For example, given a 63.1 cm3 volume and a 66.5 gram catalyst bed, a change from 290 ms (13.1 SLPM flow) to 70 ms increases the syngas flow to the reactor to 54.1 SLPM. The composition of the synthesis gas and the details of the transition are shown in Table 2 below. The temperature is raised from ~ 206.6 ° C to ~ 263 ° C to maintain ~ 70% CO conversion. The result of this high heat duty - the performance of the coolant to remove heat has been tested. Table 2 shows the key data of these operating steps.

Table 2: Process data of all welding reactors for various contact times

Based on the above data, the welding reactor design of the present invention is capable of handling thermal loads four times or more than that generated in a 290 millisecond contact time condition (with average steam quality increasing from ~ 1.5% to ~ 10%) can do.

Test conditions and comparisons similar to those performed with individual steps of operation and with other devices (blazes and single channel reactors) are described below.

The welding reactor of the present invention is tested at a contact time of 210 ms from 945 to 1131 hours in the stream. Other process parameters are H _2: CO = is kept constant at 2.0, P = 350psig (2413 kPa ), ~ 16.5% diluent. The reactor temperature is increased to ~ 214.6 ° C to maintain the target CO conversion.

The welded reactors of the present invention were tested in the stream at a contact time of 150 ms from 1132 to 1182 hours. Other process parameters are H _2: CO = is kept constant at 2.0, P = 350psig (2413 kPa ), ~ 16.5% diluent. The reactor temperature is increased to ~ 221.7 ° C to maintain the target CO conversion.

The welded reactors of the present invention were tested in the stream at a contact time of 100 ms from 1205 to 1350 hours. Other process parameters are H _2: CO = is kept constant at 2.0, P = 350psig (2413 kPa ), ~ 16.5% diluent. The reactor temperature is increased to ~ 241.2 ° C to maintain the target CO conversion. During ~ 1221 - 1228 hours in the stream, the CO flow controller is damaged due to dripping of water and must be replaced due to an interlock situation.

Thereafter, the contact time is gradually lowered stepwise from 100 ms to 70 ms. Other process parameters are H _2: CO = is kept constant at 2.0, P = 350psig (2413 kPa ), ~ 16.5% diluent. The irregular propagation behavior of the contact time of 70 ms and the reactor temperature of ~ 263.1 ° C (~1542 hours in the stream) is recognized at multiple thermocouple locations on the reactor. The sudden rapid increase in temperature can be seen under certain conditions as shown in the graph of Fig. 11 (before progress is controlled by lowering the reactor temperature).

The temperature shown in Fig. 22 is measured on the outer metal wall of the reactor on the opposite face of the 1.27 cm (0.5 inch) thick support plate in contact with the outermost coolant channel. The metal temperature is less than 10 ° C, but the rise in temperature on the catalyst bed is expected to be greater than 50 ° C.

During this operation, the high water solubility of the welding reactor of the present invention also appears. The wax material from the contact times of 290 ms, 210 ms and 150 ms is analyzed to yield an alpha number. Device performance is summarized in Table 3. Table 3 shows the high acceptability of the inventive welding reactor with an alpha value of 0.89 or more for the product wax. For contact times greater than 210 milliseconds, the alpha value is 0.91 or greater. Alpha is graded and defined as is known by those skilled in the art of Fischer Tropsch.

Contact time 290 ms 210 ms 150 ms Temperature 206.5 ° C 214.6 ° C 221.7 DEG C Performance CO conversion 74.1% 72.2% 71.1% CH4 selectivity 8.7% 9.5% 14.4% C5 + Productivity ~ 0.7 GPD ~ 0.95 GPD ~ 1.1 GPD kg C5 + / Lcat / hour 1.24 1.75 1.99 Alpha 0.91 0.91 0.89

steam quality /part Boiling  Stability determination

In the irradiating portion from 1662 to 1783 hours for the stream, the coolant flow was lowered from 0.5 LPM to increase the boiling range at the same heat duty, and [H2: CO = 2, 16.5% dilution, 350 psig (2413 kPa) To keep the other operating parameters constant from the pressure) to achieve high outlet steam quality. At ~1, 172 hours in the stream, at 0.2 LPM of flow, the flowmeter reaches the lowest readable limit, and the quantity is not further lowered. The average stream quality during this phase of testing increases to ~ 15% as shown in Table 4.

The overall weld reactor performance of the present invention at various outlet steam qualities contact
time
[ms] Coolant
Flow
[LPM] CO
conversion
[%} CH4
Selectivity
[%] Average exit
Steam quality
[%] 278 0.41 73.3 8.2 1.5 150 0.53 71.0 14.3 2.9 70 0.50 69.6 39.9 10.5 90 0.37 57.8 34.1 8.9 90 0.22 57.8 34.1 14.8

The pressure on the laser welding subassembly

During FT reactor operation, the laser welding subassembly moves from compression to tension on subassemblies because fluid pressure inside the coolant subassembly increases with time for the stream. In particular, the section of the coolant subassembly adjacent to the process layer is smaller than the entire coolant subassembly. For the entire weld reactor described in this example, approximately 60% of the subassembly is adjacent to the finned process layer and changes in compression and tension occur. A large reactor with a 24 "x 24" (60 cm x 60 cm) subassembly is greater than 80% of the subassembly surface area that transitions between compression and tension because the boiling temperature varies with pressure. The reaction temperature is typically increased for Fischer Tropsch as the catalyst is deactivated by the build up of the wax. Typical starting temperatures are between 200 and 210 degrees Celsius, and the boiling temperatures from the steam curve are between 210 psig and 260 psig (1448 and 1793 kPa). The process feed pressure is typically between 250 psig and 450 psig (1724 and 3101 kPa). The temperature is raised by increasing the pressure on the coolant side. At 220 占 폚, the steam pressure in the boiling (the preferred method for removing the Fischer Tropsch reaction) is approximately 320 psig (2206 kPa). At 230 캜, the steam pressure is nearly 380 psig (2620 kPa). In the experiment described in this example, the process pressure is raised to 250 ° C or higher, and the steam pressure is approximately 560 psig (3861 kPa) and above the process reaction pressure. At start-up, the laser welds are compressed from high pressure on the pin or process side. At the highest temperature, the laser welds are compressed and are as high as 332 psig (2289 kPa) pressure on the coolant side of the process side. Before and after the regeneration, the reaction temperature is lowered to 220 DEG C or lower, and the laser welded portion is returned to the compressed state without being put on the tension. The laser welding subassembly operates robustly in both compression and tension and returns to compression within a time lapse of more than 1000 hours for the stream. Further, not only is the laser welded part robust to operating and returning to either compression or tension, but also thermal contact of the fin with the catalyst is not affected by pressure changes in the device. An important parameter for maintaining the good performance of the reactors of the present invention is that the catalyst rod density is within 2% of the theoretical PABD (pack average bed density) as determined externally by the ASTM method for particulate material, Lt; RTI ID = 0.0 > 1%. ≪ / RTI > The cooling flow pressure drop test is used to compare the actual pressure drop predicted from the Ergun equation. If the pressure drop is not the expected 5% (2% preferred) from the Ergon equation, the dead is poorly packed. If the bed is poorly packed, the deleterious effects of flow channeling of the reactor are likely to occur, and a change between compression and tension for the process and cooling fluid may cause an undesirable increase in force on the catalyst particles, such that friction or abrasion may occur. When the catalyst particles are fractured, the resulting particulates tend to cause high pressure drops in some channels over the other and cause flow imbalance, high temperature spot or premature thermal runaway.

To summarize, the performance of a pre-welded reactor has been demonstrated and indicates that brazing and / or diffusion bonding is not required. Thermal contact is possible by a pre-weld manufacturing technique which provides good results.

The initial coolant pressure above the process pressure is used to prevent wave form deformation and popping of welds between coolant channels based on the explosion test of similar sized laser welds and the size of the laser welds to the reactor of the present invention. Is limited to 50 psi (345 kPa). During the process, this overpressure increases to ~ 332 psi (2289 kPa) and no external visual deformation is perceived. After the test, it was confirmed that the decomposition of the reactor did not cause the interior of the reactor to be compressed, deformed, or undesirably changed. This surprising result is that the catalytic rod pin is provided as a structural support for small laser welds with a 0.002 inch (0.005 cm) width. The laser welds on the can can be 0.006 inches wide (0.0150 cm), for example, to allow a sufficiently high pressure action on the coolant without requiring support from the catalyst rod process channel. Thus, in a preferred embodiment, the apparatus and method of the present invention can be used for laser welding having a width of at least 0.015 cm.

Example 3 Large sheet laser welding

Three types of parts are tested for the reactors of the present invention to illustrate the impact of dissimilar portions for a laser welding subassembly. The theoretically thin topsheet produces less distortion because of the low pressure energy to form the weld. Surprisingly, it has been found that a thin top sheet imparts more distortion. A thick top sheet is preferred.

● Top sheet welded to 0.020 "(0.05 cm) wall, 0.010" (0.025 cm) wall, 0.005 "(0.025 cm) walled coolant channel shim

• The initial weld is performed by intermittent laser welding (eg every fifth row) to assist in securing the parts. Intermittent laser welding along the weld length dimension is added to every fifth row before the long seam weld line is formed. Intermittent stitching is approximately 2 to 4 cm long and is separated by an unwelded section of 5 to 20 cm in length.

• Afterwards, the top plate is removed and a sufficient length of weld is performed.

• Power settings and focus are required to be adjusted for two thin wall scenarios.

○ 60% power setting for 0.010 "(0.025 cm) wall

○ 50% power setting for 0.005 "(0.0125 cm) wall

The initial idea was that distortion was reduced as wall thickness and power decreased, but the opposite phenomenon occurred. 0.020 " (0.05 cm) wall subassembly was measured and ~ 2.750 " (6.875 cm) distortion appeared.

Continuous testing for 0.010 " (0.025 cm) and 0.005 " (0.0125 cm) walls shows the following number of stress variations.

The 0.010 "(0.025 cm) subassembly measures 2.906" (7.265 cm)

The 0.005 "(0.0125 cm) subassembly measures 2.961" (7.4025 cm)

It is not surprising that, in the subassembly tested, there is a weld of 161 24 "(60 cm) in length along a 24" (60 cm) width of the assembly and that stress can be seen in these components, It is not the normal longitudinal direction and transient stress. There is a challenge to mitigate stresses during the welding process or to reduce and eliminate stresses after the subassembly process.

Preferably, the individual topsheet has a thickness of at least 0.04 cm, in some embodiments in the range of 0.04 to 0.2 cm, more preferably 0.05 cm to 0.1 cm. &Quot; Top sheet " refers to a sheet or sheet comprising a plurality of sheets or channels containing channels or other voids, wherein the top sheet seals a channel or cavity in the height direction and completes the subassembly.

Preliminary cambering test for large part

One way to alleviate the stresses in the welds was to test the operation by means of a pre-cambering (pre-bending) component.

● Process

○ As a unit, preliminary camber coolant and coolant channel shims

Load the fixture onto the laser weld table and perform setup and laser alignment

○ Perform laser stitching program

○ On the laser welding table,

○ Performing a sufficient welding program

The pre-cambering component is made using a fixture configured as a clamp and fixture to allow controlled, consistent bendability of the component (see Figure 12). The sheet is bent in a direction perpendicular to the channel length. The two sheets are pre-cambered and then welded with this technique. The final subassembly is visually compared to a similarly fabricated subassembly without pre-cambering.

Stress relief at temperatures below 400 ° C is an acceptable practice, but only the appropriate stresses are eventually mitigated. One hour at 870 ° C typically relaxes about 85% of the residual stress. However, stress relaxation at these temperatures can precipitate grain boundary carbides, seriously damaging the corrosion resistance in many media. To avoid this effect, it is recommended to use stabilized stainless steel (321 or 347) or extremely low carbon (304L or 316L), especially when long stress relief is required. A large part of the stress relaxation is attempted at a high temperature of 400 ° C and 1100 ° C. In both cases, limited success was observed. The appropriate reduction in deformation was observed ~ 40% of the initial distortion as measured by the distance of one edge over the flattened portion.

• Annealing (often referred to as liquid treatment) not only recrystallizes the work-hardened grain, but also retreats the chromium carbide (precipitated at the gray boundary in the sensitive steel) to the austenite solution. In addition, the treatment uniforms the dendritic weld metal structure, and all remaining stresses are relieved from the cold work. Although some forms may be annealed at a closely controlled temperature as low as 10 < 10 > C when fine grain size is important, the annealing temperature is typically above 1040 < 0 > C. Time at temperature often keeps surface scaling short to minimize or control grain growth that can be induced to " orange peel " upon formation.

Example 4. Pin Impact in Welded Reactor

A waveform with a press-fit pin with a high aspect ratio is not straight, but has some bending or bending on the pin. After being compressed in contact with the two solid strips on either side of the pin, the pin is further bent or bent. In the welding reactor of the present invention, unlike brazing or bonding, the fins are pressed against the adjacent surfaces and become contact resistance or thermal resistance at the contact points of the adjacent surfaces with the fins. In a heat or endothermic reaction, the heat of reaction is transferred between the pin and the adjacent surface. To improve thermal contact with the press fit, the fin is larger than the support edge strip. The pin is pressed and contacted by the outer rod, thereby being bent or buckled. As the pin is bent, the original strength is hard to maintain.

A 0.006 "(0.015 cm) thick Cu110, 0.256" (0.64 cm) high pressed pit pin is pressed against an edge strip of 0.25 "(0.63 cm) height. The pin is eccentric when pressed against an adjacent heat transfer wall. In the picture of Figure 13, the horizontal line shows the pin placed on top of the laser welding line sealing the cross flow heat exchange channel.

If the press-fit pin described in Example 1 is surprisingly well performed, the filling of small gaps (measured between about 5 and 150 microns) with the Fischer Tropsch liquid and / or hydrogen produced during the reaction and / This is equivalent to the performance of a braze reactor that does not have a laser welded ridge that isolates the fins from the heat transfer wall as a result of a combination of high thermal conductivity of the copper fin that transfers more heat to the heat transfer wall, including assisted by the conduction. The thermal conductivity of the liquid oil and hydrogen is substantially higher than most gases, thus reducing the effect of poor contact resistance between the press-fit pin and the heat transfer wall. In addition, the advantage of using copper with very high thermal conductivity allows efficient axial conduction to transfer heat to the heat transfer interface without building hot spots or creating poor side reactions or making them worse.

A load of 2600 psi (17,926 kPa) is applied to the stack to contact the stack. A preferred range of loads may range from 500 psi (3447 kPa) to 500,000 psi (3,447, 000 kPa), depending on the height of the pins, the material to be peened, the thickness of the pins and the eccentricity of the initiation pin. After the primary TIG welding along the compression and side bars or edge strips is completed, the pin contact is made with the heat transfer subassembly.

Example 5. Large laser welding subassembly linear density of weld per unit area of subassembly

This example illustrates a 24 " x 24 " (60 cm x 60 cm) laser welding subassembly having a laser welding line running along a length of 0.6 m in the direction of cooling flow. There are 161 coolant channels and 162 laser welds on these panels to seal the device between coolant channels and provide structural support to prevent deformation of the equipment during operation under pressure.

In this example, there is a linear weld density of 270-m per square meter of surface area or a linear weld of 97.2 m at 0.6 mx 0.6 m. Alternatively, the linear density is described as 2.7 cm / cm ² at 3600 cm ² size in this embodiment. In this embodiment, the linear density of the weld is large or small, or may be more, preferably 0.1 cm / cm ² to 10 cm / cm ² than 2.7 cm / cm2.

Example 6: Laser welding registration

Figure 14 shows a laser welding line that is joined to the top of the rib between parallel adjacent coolant channels formed in the bottom plate. The laser weld joins the bottom channel plate to the top plate. Laser welding is applied along the top of the ribs formed in the bottom or channel plate. In this example, the ribs are 0.037 "(0.093 cm) wide and the laser weld width can vary from 0.002" (0.005 cm) to 0.01 "(0.025 cm). It can be on either side or anywhere that follows.

Example 7: Fabrication of large FT reactors

The pre-welded Fischer Tropsch reactor core is preferentially built as a dual layer assembly of other coolant and process parts. A feature of this design is that the coolant subassembly is welded in a manner that maintains mechanical integrity as a stand-alone unit. One process for creating such a subassembly is by using laser welding to join the upper solid shim to the featured shim by placing welds between each of the channels and proceeding parallel to the ribs. The required mechanical integrity has proven to be achieved with this approach. The second parameter to be described is partial flatness. Welding of two thin sheets in this manner may cause significant subassembly deformation due to shrinkage of the welded stainless steel and the welded material. Subassembly deformation requires additional effort to add complexity to re-planarize the piece or to apply it to a lamination process that handles deformed pieces.

One method that has been found to minimize warping is to limit the overall size of the subassembly. While maintaining a level of acceptable flatness by using subassemblies that combine 24 "(60 cm) long welds between channels limited by a width of approximately 6" (15 cm), they have a width of 24 "(60 cm) (I.e., the entire portion is 6 " x 24 ", 24 " x 24 ", (15 cm x 60 cm vs 60 cm x 60 cm) Subassemblies can be stitch welded together to maintain reasonable planarity. In this manner, a 24 "x 24" (60 cm x 60 cm) flat subassembly is 24 "x 24" (60 cm x 60 cm) welded from a 24 "x 24" (60 cm x 60 cm) Can be erected by stitch-welding four 6 "x 24" (15 cm x 60 cm) subassemblies that are substantially planar than subassemblies.

Another useful feature that can be built into the weld subassembly is to precisely check for leaks and mechanical integrity prior to use to build up the entire FTR weld stack. This can be accomplished by slightly increasing the initial subassembly, adding a port to which pressure is applied, and initially sealing the end of the subassembly as part of the laser welding process. These pressure tests may be hydrostatic or pneumatic. After the individual pieces have been welded and fitted, the subassembly is cut to size precisely 6 "x 24" (15 cm x 60 cm) and to form one 24 "x 24" (60 cm x 60 cm) The four 6 " (15 cm) subassemblies of the stitches weld together and open the flow end of the channel.

The final subassembly is then interleaved between the processes to form the main reactor core. The lamination process first lowers the 2 "(5 cm) thick clamp plate and the 1" (2.5 cm) thick end plate, and then proceeds to the alternating coolant and process layers. The stack is terminated by disposing the final coolant subassembly, top plate, and top clamp plate. Pressure is applied to the stack to pre-compress the copper waveform so that all components are in metal-to-metal contact. The applied pressure may range from 20 psi (138 kPa) to 500,000 psi (3447000 kPa), preferably from 20 to 20,000 psi (138 kPa to 138000 kPa), more preferably from 20 psi to 5,000 psi To 34474 kPa). Thereafter, the stack is clamped in place using the clamping system before releasing the applied pressure. The clamping system maintains the stack in a pressurized state so that core welding can occur.

Core welding consists mainly of three steps: adding two weld seams to both process surfaces, adding sealing welds to both cooling surfaces, and applying two end plate seals to each process face to prevent bypass And adding welding. The reactor is left in a clamped state during each welding step to ensure the best possible thermal contact within the corners. It should be noted that each of the three welding steps performs its own purpose. The reinforcement welds that occur on the process side are applied first and provide sufficient mechanical strength to allow the core to be easily manipulated (raised, rotated or otherwise directed) during the subsequent fabrication steps as well as the other two weld steps on the entire stack. Sealing welding occurring on two cooling liquid surfaces is the primary welding that protects the internal cross leakage in the reactor (process to coolant or vice versa). End plate welding is used to seal the outermost coolant subassembly to the top and bottom end plates. Even if the cooling liquid surface is sealed to avoid cross leakage, it must also be sealed to the process surface to bypass the catalyst bed to avoid process gas transfer between these parts. It should be noted that the catalyst is loaded into the reactor between the process pins after the pin assembly.

Before the core is welded to the assembly, the reactor should be practically free of leaks. The welding reactor is not yet prepared to withstand any significant internal pressure, so that the bolts of the clamping mechanism can be used to provide support. Once the reactor core is fitted, the coolant header welds proceed to have their positions. Once these steps are performed, they can pass coolant side flow tests if they are desired or justified. The corresponding coolant footer is then welded in place. All of the coolant manifolds are suitable for operation and provide a portion of the base of the reactor's external support system.

Because the pre-weld is made at the periphery of the core, with the exception of the coolant subassembly, the reactor can not withstand the intrinsic process pressures without deflecting the top and bottom end plates due to the pressure-loaded rod, The operating state can not be achieved. To impart mechanical integrity to the reactor, the system of the external support (exoskeleton) is welded around the reactor core. This support is designed to balance the internal process pressures to control any pressure-induced deflection of the top and bottom plates to an acceptable level. The outer support acts as a stiffener that is welded across both the top and bottom end plates of the reactor and then welded from top to bottom. It balances any load created by internal compression and prevents any other deformation that occurs. (See Fig. 15). Preferably, each reinforcing element is at least three times thicker, preferably at least five times greater than the thickness. (Larger is the thickness of the reinforcing element in the stacking direction. In some embodiments, the thickness and height as well as the spacing between sets of supports are determined based on a balanced process load. In one example, the spacing between sets of supports is produced from a 0.75 inch thick stainless steel plate extending approximately 8 " high across the top and bottom end plates with approximately 3 " spacing. As a final conformity determining step, the core may have a temporary manifold welded on the process side and subjected to high pressure testing to ensure that the reactor meets design criteria. After this step, the process manifold is removed and the core can be welded in place after the final process manifold has been reserved for catalyst loading

Figure 16 shows an example in which four 6 " x 24 " (15 cm x 60 cm) subassemblies are arranged side by side and welded at several locations to bond with a 24 "x 24" (60 cm x 60 cm) assembly. Spot welding is preferably bonded to subassemblies in this manner because continuous welding results in more deformations.

Example 8

The apparatus of Example 8 is a welding reactor or apparatus that provides cross flow heat transfer between two fluid streams. Although other flow arrangements may be used, the particular illustrative example is cross flow. The outer support is welded to the outside of the device to allow the pressure of the inner passageway of the device relative to the low pressure outside environment without compromising the integrity or restraint of the device (see Figure 17). Or the array allows the device to withstand high pressures that are different from the external environment. The device consists of 304L stainless steel with an external support. The use of the exoskeleton permits the application of heat exchange without the use of an external pressure vessel or the operation of high pressure applications or Fischer Tropsch reactions and other reaction weld reactors. The device is in a tensioned state different from compression due to the presence of the external pressure of the container having the high-pressure fluid surrounding the reactor or the apparatus of the present invention.

The 61 cm x 61 cm x 6.5 cm device core of Example 8 consists of a layer welded around the periphery as described in the appended application. The outer support is about 14 cm in width near the end and about 17 cm in width in the area adjacent to the 61 cm x 61 cm side, 1.9 cm thick and 105 cm long. The support is 10.2 cm apart (center-to-center) with a 1.9 cm thick cross member positioned between the supports along each device (so that the two rows of cross members are spaced about 60 cm apart). The welding between the outer support and the cross member is a full penetration bevel welding.

A hydrostatic test is first performed on the process stream flow circuit. The procedure used (shown graphically in Figure 18) is as follows.

1. Measure baseline leaks during pressure test from nitrogen at about 690 kPa (psig)

2. [In the example where Lab Alliance HPLC "spare pump" is used] The pump is used to charge the device with water

3. Pump used for pressure rise from atmospheric pressure (ie, <450 kPa) to ~ 3300 kPa (464 psig) at a rate of ~ 300-400 kPa / min

4. Pressure drop to ~ 3000 kPa (420 psig) at a rate of ~ 50-100 kPa / min

5. Pumps used for pressure rises from ~ 3000 kPa (420 psig) to ~ 3700 kPa (522 psig) at speeds of ~ 50-100 kPa / min.

6. Pump used to increase pressure from ~ 3700 kPa to> 6000 kPa (855 psig) at a rate of ~ 100 to 150 kPa / min.

7. Lowering the pressure to ~ 5300 kPa (754 psig) at a rate of ~ 250-300 kPa / min

8. Continue to descend the pressure until the ambient conditions are reached and water is drained from the unit

9. Repeat step 1

A second hydrostatic test is then performed on the cooling fluid stream flow circuit. The procedure used (shown graphically in Figure 19) is as follows.

10. [In the example where Lab Alliance HPLC "spare pump" is used] Charge the device with water using a pump

11. Pump used for pressure rise from atmospheric pressure (ie <250 kPa) to ~ 3500 kPa (495 psig) at a rate of ~ 2000-2500 kPa / min

12. Pump used to increase pressure t from ~ 3500 kPa to> 6000 kPa (855 psig) at speeds of ~ 800-900 kPa / min.

13. A pressure drop of less than 5200 kPa (740 psig) at a rate of ~ 400 kPa / min

14. Continue to lower the pressure until the ambient conditions are reached and water is drained from the unit

15. Compare the leakage of approximately 690 kPa (100 psig) with the baseline pressure test leakage

Hydrostatic testing of the coolant and process circuit is performed using the above protocol. The device has no indication of mechanical damage during hydrostatic testing. The leakage from the coolant circuit to the process circuit before and after this hydrostatic test, measured with a pressure drop of more than 15 minutes at 690 kPa (100 psig) at the coolant circuit, was 0.6 kPa (0.09 psig) and 21 kPa )to be. The device is then repaired by welding and the amount of leakage from the coolant to the process circuit is measured as a pressure drop of 2.2 kPa (0.32 psi) for at least 15 minutes at an initial pressure of 690 kPa (100 psig). Repair welding is performed using a fiber laser. Alternatively, TIG, MIG or other conventional welding methods may be used.

The weld reactor core is assembled from the laser welding coolant subassembly and welded to the final reactor. Thereafter, the reactor is continuously housed in an exoskeleton that performs high-pressure service. For the Fischer-Tropsch reactor, the hydrostatic pressure test is 855 psig (5895 kPa). Alternatively, high or low fluid dynamic pressure testing is used for Fischer Tropsch services according to the final desired operating conditions. The hydrostatic test described in this example makes the reactor suitable for service at peak design pressures of 562 psig (3875 kPa) and peak design temperatures of 250 ° C. The operating temperature and pressure are lower than the peak design pressure to allow margins of operational stability. For a preferably high operating pressure on the Fischer-Tropsch reactor, the spacing of the outer support as shown in Fig. 17 is reduced and the additional support is added to accommodate the higher operating pressure of the reactor.

When operated at high temperature or pressure, other reactions are possible with the exoskeleton of the present invention, but more densely spaced additional supports are required. Alternatively, as the desired process design pressure or temperature is lowered, a welding support bar of a leaner density is used. The welding reactor of the present invention can be operated at high temperature and high pressure without being placed in a pressure suppression vessel or without a brazed or bonded reactor core.

The exoskeleton of the present invention allows the reactor or device to pass hydrostatic testing and the high and low external pressures of the reactor maintain mechanical integrity.

Example 9 - Leak test

Pre-welder leak check test

summary

The full weld FT unit contains 100 psig (690 kPa) independently on both the process and the coolant, and the leak is aerodynamically checked. In some embodiments, the leak check pressure may be as high as 200 psig (1380 kPa), or 500 psig (3450 kPa), and in one embodiment as high as 1000 psig (6900 kPa). The pressure drop over time is recorded to determine if device leaks and interfaces have been snooped (leak-tested) to identify any leak location. To allow compression on either side, the gasket head and footer are secured to the device via all threaded clamping setups. These devices are compressed in increments of 10-20 psig (90-180 kPa), stop after each increment, check the pressure drop rate, identify any leaks in the device, and check all gaskets and fittings for leaks Check.

Example 10. An ultra-tall pin available by the reactor of the present invention for Fischer-Tropsch and other chemistries.

A pre-weld reactor in which a process layer (including a fin structure) is positioned adjacent to a coolant subassembly to form a final device assembly allows the use of conventional process layers including those with ultra-toll pins. Ultra-tin fins have problems such as sagging and deformation during brazing or bonding, and if the structural support is small or available during the bonding or framing process, it will have a final structure that requires fixtures or straight holders before use. Excessive deformation from ultra-tin pins after brazing or bonding (about 0.5 in. Or 1.25 cm or more) will render the device unusable.

Case A: Pin height 0.0225 "(0.5625)

The Fischer-Tropsch reaction is performed in a microchannel. The microchannel reactor includes a plurality of parallel process panels packed with an FT catalyst. The heat of reaction is removed by the coolant channel between the process channels, and the coolant is water. The elimination of the reaction heat generated inside the catalyst bed is improved by the buried fin structure. The continuous fins inside the process channel form a series of parallel flow paths. In this example, the pin is made of copper 110. The process channel height is 0.225 "(0.5625 cm) and the length is 23" (57.5 cm). The pins have the same length and the spacing between adjacent pin ribs is 0.04 " (0.1 cm) and the pin thickness is 0.006 " (0.015 cm). There is no contact resistance between the pin and the process channel wall. The process channel wall thickness is 0.02 " (0.05 cm), and is made of stainless steel.

The process channel includes a predetermined amount of the Co catalyst and is described in the first example. The void fraction of the catalyst bed is approximately 0.4 and the effective thermal conductivity is approximately 0.3 W / m-K. The catalyst loading is 1060 kg / m &lt; 3 &gt;.

The complex TF reaction is modeled as a networked simplified reaction with a six-volume reaction (see FIG. 5). The parameters in the velocity representation (Table 6) are pumped using the catalyst test data in a lab scale TF reactor.

Table 5 FT Reaction and Exercise

Table 6 Parameters in speed equation

This is the operating condition used in the FT reactor model.

● Temperature on the process channel wall: 230 C

● Starting pressure of catalyst bed: 412 psig (2840 kPa)

● Supplied H2 / CO ratio: 2: 1

● Supplied nitrogen dilution rate: 31.3% (vol)

• The process feed is preheated to the same degree as the channel wall.

● The feed flow rate is 0.097 "(0.1 cm) x 0.225" (0.563 cm) x 23 "(57.5 cm) one unit catalyst packing over: 1197 SCCM The contact time calculated based on catalyst volume is 0.17 second.

Figure 20 shows the expected catalyst bed temperature as a function of reactor length. The temperature is sampled along the center of the catalyst bed such that the peak of the curve is represented by the maximum temperature of the catalyst bed. In this case, it is 239C and is located a short distance from the beginning of the catalyst bed. CO conversion is expected to be 76.0% and methane selectivity is 15.3%.

Case B: Pin height 0.5 "(1.25 cm)

The Fischer-Tropsch reaction is performed in a microchannel reactor. The reactor configuration is similar to the reactor in case A. The only difference is that the process channel and copper fin height is 0.5 "(1.25 cm). For normal 0.5" (1.25 cm) pin heights, the starting pin height is normal 0.5 "(1.25 cm) Or between 0.051 "(1.253 cm) and 0.052" (0.13 cm), preferably between 0.504 "(1.26 cm) and 0.510" (1.275 cm), located after the p-strip.

In this example, the operating conditions are the same, except that the feed flow rate increases with the overall catalyst loading volume to maintain the same reaction contact time of 0.17 seconds. The flow rate is 2661 SCCM.

The same catalyst and dynamics are used in this example. The characteristics of the catalyst bed are the same as those used in case A.

Figure 21 shows the expected catalyst bed temperature as a function of reactor length. The temperature is sampled along the center of the catalyst bed such that the peak of the curve is represented by the maximum temperature of the catalyst bed. In this case, it is 246 ° C and is located a short distance from the start of the catalyst bed. CO conversion is expected at 81.9% and methane selectivity is 17.0%. It is anticipated that the operating temperature will be slightly reduced from 226 캜 to 229 캜 when driven by the steam side pressure so that the CO conversion is less than about 76% and the corresponding selectivity measurement is expected to be slightly less than the expected 17%. Overall, the reactor is expected to be thermally controlled to a 0.5 "(1.25 cm) fin height. The pre-weld reactor of the present invention is expected to be able to accommodate a copper pin height of 0.5" (1.25 cm) for a featured reactor . This translates to ultra-tall pins [0.25 "(0.625 cm) or more) inability to accommodate any of the stainless steels for featuretrops or braze reactors. Ultratolfine from the braze reactor is a high temperature and high load brazing process The use of a pre-welded reactor platform gives more advantages to the toll pin and the ultra-toll pin Fischer Tropsch reactor.

Case C: Pin Height 1 "(2.5 cm)

The Fischer-Tropsch reaction is performed in a microchannel reactor. The reactor configuration is similar to the reactor in case A. The only difference is that the process channel and copper fin height is 1.0 "(2.5 cm).

In this example, the operating conditions are the same, except that the feed flow rate increases with the overall catalyst loading volume to maintain the same reaction contact time of 0.17 seconds. The flow rate is 5321 SCCM.

The same catalyst and dynamics are used in this example. The characteristics of the catalyst bed are the same as those used in case A.

Figure 22 shows the expected catalyst bed temperature as a function of reactor length. The temperature is sampled along the center of the catalyst bed such that the peak of the curve is represented by the maximum temperature of the catalyst bed. In this case, it is 600 C or more. Since the amount of catalyst is almost four times that in the TF reactor at A, the total heat of reaction through each pin rib is increased by the same scaling factor. At this level of reaction heat, a low heat transfer resistance inside the copper rib is important. This can be seen from the critical pin temperature fluctuations from the center to the edge. By the high temperature in a large part of the catalyst bed, the CO conversion is expected to be greater than 90% and the methane selectivity is significantly higher than in the case of the short fin height. In this type of heat dissipation, the TF catalyst is expected to be rapidly deactivated and heat dissipation will continue to move downstream at many similar axial positions on the burning cigar. After burning out the FT catalyst, the overall conversion is low (less than 40% per pass at similar flow, temperature and pressure conditions) and the corresponding methane is high (greater than 10%).

It is believed that the above example completes the thermal contact of the fin to the wall and completes the flow dispersion on the side of the cooling fluid with no block or low flow cooling fluid channel. Large scale reactors actually produced with the use of particulate FT catalysts having a pin height of preferably less than 0.5 inch (1.3 cm) to accommodate operational flaws or manufacturing defects due to potential fouling on the coolant side during operation do.

.

Claims (21)

A laminated microchannel assembly having a first metal sheet and a second metal sheet,
Each sheet having a length and a width, the cross-section defined by the respective sheet length X sheet width being greater than 100 cm2,
Wherein the first and second sheets are flat,
Said first sheet comprising an array of parallel microchannels, said microchannels being spaced from each other by barrier walls,
The first sheet and the second sheet being adjacent,
And a weld that follows the length of the barrier wall and bonds the first sheet to the second sheet. The microchannel assembly of claim 1, wherein the cross-sectional area defined by the sheet length X sheet width is 500 cm 2 or greater. 2. The microchannel assembly of claim 1, wherein the weld is continuous. The microchannel assembly of claim 1, wherein the weld on the second sheet has a linear density of at least 2.7 cm / cm2. The microchannel assembly of claim 1, wherein the weld is a laser weld having a width of at least 0.015 cm. A method of manufacturing a laminated device,
Providing a first subassembly or first sheet and a second subassembly or second sheet,
Welding an edge of a first assembly to an edge of a second assembly to form a subassembly layer combined by a welding edge of the first and second sheets to form a welded single sheet; Laminating a welded single sheet to one or more layers or sheets; and joining the laminated layers or sheets to form a laminated device,
The first subassembly or first sheet having a first parallel array of channels,
Said second subassembly or second sheet having a second parallel array of channels,
Wherein there is no intersection between the first subassembly or first sheet and the second subassembly or second sheet. 7. The method of claim 6, wherein the first and second parallel arrays of channels are capable of sharing a common header or footer. 7. The method of claim 6, wherein the first assembly is welded to the second assembly by spot welding. 7. The method of claim 6 wherein the single sheet or subassembly is cut together into a plurality of pieces and welded together to form a continuous assembly. 7. The method of claim 6, wherein the first subassembly or first sheet is the same size as the second subassembly or second sheet. 7. The method of claim 6, further comprising planarizing the first subassembly or first sheet prior to welding the edges together. A laminate-like device produced by the method of claim 6. delete delete delete delete Wherein each sheet has a cross-sectional area defined by the sheet length X sheet width greater than 100 cm2 and having a linear density of laser welded joints between 0.05 and 20 cm / cm &lt; 2 &gt; An assembly is bonded to the top plate to form an assembly over a sheet surface section in the microchannel device, the section comprising at least 50% of the adjacent area of the major surface, the major surface being a surface formed by laminated and bonded sheets, Wherein the sheets are metal sheets. 18. The micro-channel device of claim 17, wherein the cross-sectional area defined by the sheet length X sheet width is greater than 500 cm &lt; 2 &gt;, and wherein a weld assembly having a linear density of laser welded joints between 0.05 and 20 cm / Wherein the section comprises 100% of the adjacent area of the major surface. A method of forming a laminate-type assembly,
Comprising the steps of: welding an upper sheet to a bottom sheet to form an array of parallel microchannels disposed between an upper surface of the upper sheet and a bottom surface of the bottom sheet, wherein a seal is provided between the channels in the array of parallel microchannels Welding is used to form,
Each sheet having a length and a width,
Wherein the cross-sectional area defined by the respective sheet length X sheet width is greater than 100 cm &
Wherein the parallel microchannels are spaced from each other by a barrier wall. 20. The method of claim 19, wherein the method of joining the sheet to the assembly comprises a laser welding step for sealing between two adjacent internal flow microchannels. delete

Images